Excess Spread is our weekly newsletter, covering trends, deals and more in structured credit and ABS. Find out more about 9fin for structured credit.

So much worse

There is still much to come out on the fate of Stenn International, the trade receivables ‘fintech’ formerly valued at $900m, and now little more than a smoking ruin, with around £2m of cash to cover £90m or so in liabilities, according to administrators Interpath.

But we’re getting more details as the dust settles.

According to court documents obtained by 9fin, “hundreds of millions of dollars” of receivables “do not seem to represent real debts”, as determined by an investigation conducted by HSBC Innovation Bank, the corporate lender to Stenn. The court transcript cites four examples coming to over $220m, and note “there are likely to be other examples”.

There was $978m of receivables in Stenn Asset Funding just before the administration, and senior lenders were protected by $100m of mezz plus Stenn’s equity; despite last week’s optimism, it looks like substantial principal writedowns for the senior are indeed on the cards.

Bloomberg had reported that some of the firms on Stenn’s client list denied doing business with the firm, but the administration documents lay out in more detail what went wrong.

According to a statement by Stephen Robins KC, acting for HSBC Innovation Bank, “The invoices are said to be debts owed by large blue-chip companies. But the payments are made by companies with similar names which, on further investigation, have no connection with the blue-chip companies named on the invoices, and often those companies making the payments are incorporated in jurisdictions such as Serbia, Hong Kong and Thailand. Often, as you will have seen, they share the same directors even though they are ostensibly unconnected with each other, and often those directors or other individuals are the ultimate beneficial owners of the paying companies, which, as I said, seem to have nothing to do with the blue-chip companies named on the invoices and you have seen numerous examples.”

Similar names!

It’s astonishing how well this apparently unsophisticated technique can work for nefarious purposes.

The gigantic 1MDB fraud perpetrated by Jho Low (read the superb Billion Dollar Whale for a gripping and deeply reported account) turned partly on vehicles incorporated by Low called Vista Equity and Blackrock Commodities (Global), which have nothing whatsoever to do with legitimate top tier financial institutions Vista Equity Partners and BlackRock, other than the name. When proceeds from the fraudulent 1MDB bonds were sent to these accounts, he simply stole the money.

The money hasn’t vanished at Stenn, as it did at 1MDB — it’s just that the receivables weren’t real debts and the companies that owed them weren’t the real companies.

But it’s still a situation which has rocked the trade finance world.

Receivables fraud is very much a known problem (invoices are just PDFs!) and part of Stenn’s USP to funders and investors was sophisticated technology to deal with precisely this issue — to verify that two parties were indeed shipping goods to each other, to comb shipping data and credit files to match up entities and names, as well as more traditional manual verification.

While the details above help understand the situation, they raise another set of questions.

Were Stenn’s systems bypassed, or were they never quite up to the job? How sophisticated was the fraud in question?

More importantly, is the classic cui bono? (who benefits, for the philistines at the back). For what purpose was the fraud done?

There are various forms of receivables chicanery, which leave different tracks. The classic method is simply to sell fake receivables to your finance counterparty, and then steal the money. This is more or less what happened in FCT Smart Tréso, which we discussed here. Greensill financed receivables that explicitly did not exist — “future receivables” — effectively turning secured financing into unsecured. Greensill was therefore taking on far more risk than it appeared to be doing, and benefited through fees and related party interests.

In the Stenn case, the money seems to have been paid — the securitisation didn’t grind to a halt over poor performance, which one would expect if hundreds of millions had been disappearing. It was locked up only when HSBC’s investigation turned up its results, immediately before administration, and Stenn remains the servicer of the runoff portfolio.

So the end result here is that money wasn’t stolen — it just moved from a series of higher-risk jurisdictions to other end clients, via a trade finance platform. For what purpose, we know not, but various possibilities exist.

The FT’s coverage notes that HSBC’s investigation was originally triggered by a passing reference in a US criminal indictment relating to money laundering — but Stenn was not the subject of the indictment, which referenced a Russian citizen called Feliks Medvedev.

At Stenn itself, it’s not clear how any controls were evaded, and who might be involved. The company had expanded rapidly in the past two years. Far-flung originators with ambitious revenue targets and poor oversight might be to blame (but how did they evade the controls supposed to exist?).

Stenn’s board agreed with the HSBC administration terms, but had previously been gearing up for a fight — Quinn Emmanuel had been engaged (though not yet paid) and lobbed in a request for £3m in legal fees, only to be informed that Citibank had already frozen Stenn’s accounts.

So a final question — what did the Stenn board hear between engaging Quinn and agreeing to the administration?

This matters not just because Stenn matters — it had nearly a billion dollars of receivables financed — but because of the broader sector. Non-bank trade receivables finance is a useful sector with the potential to do good in the world. It’s genuinely a useful way to fund smaller enterprises and lubricate global trade.

Any future funder of any similar platform will ask “how can you be sure it won’t be like Stenn”?

Without knowing exactly how the Stenn situation came to pass, that question is hard to answer. Traditional due diligence is useful, but it’s a poor defence against intentional fraud, and given the extensive list of top tier counterparties involved in funding Stenn, it clearly didn’t work here. Until counterparties can be sure they aren’t funding the next Stenn, they’ll be less willing to fund anyone.

Shoot me a message if you want a peek at the transcript — it’s short but potent stuff.

Should we worry about SRT repo?

I guess I could just write “no” in this section but that would be dull for all concerned.

My colleague Celeste reports that Deutsche Bank is pulling in its horns in the provision of SRT repo, so it’s a good time to look at the issues around the product. The institution in question is one of the largest SRT issuers, though, of course, banks will not finance their own SRT tranches on repo.

The regulatory worry is basically that SRT repo returns risk to the banking system which should have gone for good, into the arms of the protection-selling hedge fund.

Per an IMF report, which generated a certain amount of consternation in the industry:

“There is anecdotal evidence that banks are providing leverage for credit funds to buy credit-linked notes issued by other banks. From a financial system perspective, such structures retain substantial risk within the banking system but with lower capital coverage.”

This has always seemed a bit unfair, on various grounds. One is general equitable treatment across products. A struggling bank could do a rights issue, and sell new equity to hedge funds, financed by their prime brokers elsewhere in the banking system. This is less transparent, more levered, riskier and more procyclical than SRT repo, but nobody bats an eyelid.

You could also consider equitable treatment across securitisation formats. Financing junior cash ABS positions is generally fine and uncontroversial; why is it so much worse to finance SRT CLNs?

The other point is more structural. Banks divide their regulatory capital into credit risk and market risk. Many positions attract both kinds of capital charge, but generally what happens when a repo financing arrangement blows up is that the bank takes the collateral and resells it. It is subject to the market price movements against it, and that’s what the haircut is for. In SRT markets, these are generally 40%+, enough to take a pretty big price hit.

If you’re of a tranching inclination, you could view a 12.5% SRT financed on repo as, say, 0-6% investor, 6-12.5% financing desk, but that’s not really right.

If the SRT experiences more than 6% losses, it’s still the SRT investor bearing the losses. The bank will have ratcheted up its margin requirements long before. Even if it happens fast enough that margin didn’t move, it remains the investor that’s on risk for the full amount (though the financing desk would be eyeing the position nervously).

From the issuing bank’s perspective, presumably the one which regulators care most about, it doesn’t even really matter where the bonds have gone. The cash is already in the door, the deal is fully collateralised. The losses are covered even if the fund entity goes up in smoke.

We have (I think) only one good example of things going wrong; a reg cap hedge fund that was margin-called and had to liquidate its portfolio in 2020. The positions were sold on and absorbed (though with some griping about marks) and the fund is back in the market and keener than ever to buy. That’s what’s supposed to happen! It is regrettable when a hedge fund blows itself up, but Covid was unprecedented, and prudent regulation shouldn’t mean regulating risk out of existence. They’re grown-ups, let them trade!

The big wrinkle to this is correlation, but it’s a bit of a reach. Large corporates are the biggest SRT asset classes, and big IG companies show up in many of the most “liquid” SRT shelves. Big French companies that BNP Paribas banks may also be banked by Société Générale and Credit Agricole, and potentially turn up in all of their SRT deals.

So if you have some event which causes mass defaults in the largest investment grade companies, SRT funds with a lot of large corporate exposures might see their marks move against them, get margin-called all at once, and potentially trading desks will end up owning the bonds. Under this scenario, they’ll be trading below par, and possibly even below 60% (representing the haircut).

But presumably the event which causes mass defaults in the largest investment grade companies would already have rattled the banking system pretty badly, and I’d be amazed if the secured funding desks would be the first to fold in such an environment — the ideal hedge here would probably be tinned food and a shotgun.

Still, regulators gonna regulate, and even if that just consists of asking irritating questions, it may not be worth the hassle to stay in the business. There’s a lot of other illiquid structuring financing to do, on regular way ABS, infrastructure loans, CRE loans, private credit and more, and formats like fund-based NAV financing might achieve similar results without the pressure of a raised eyebrow from the IMF.

Liz is back

Elizabeth Finance 2018 achieved the dubious honour of being the first post-crisis European securitisation to see triple-A losses, following the sale of the shopping centres backing the Maroon loan last year.

Strains in bricks and mortar retail were already fairly apparent when the deal was structured in 2018, with multiple UK retailers having been through ‘CVA’ restructurings, which tended to fall heavily on landlords, with existing lease terms written down. Struggling department stores like House of Fraser (CVA May 2018, administration August 2018) were often the anchor tenants for shopping centres, occupying the largest footprints across multiple floors.

But things went from bad to worse, as Covid hollowed out the IRL shopping experience, and base rate hikes pushed cap rates up (and therefore valuations down). The assets in the Elizabeth Finance Maroon loan were fairly secondary, in King’s Lynn, Dunfermline and Loughborough; important to the small towns they served, but not the premium superregionals owned by the likes of Intu (also a Covid casualty) or Westfield.

The valuation of these three centres at origination was £104m. Heads of terms were agreed at £35m in the middle of last year, with contracts exchanged in October at £33.2m, and final recovery was estimated at £28.5m, after some tax requirements and a £350k rent top up, too little for the £33m triple-A to get out whole. This was exacerbated by structural weaknesses in the deal; previous paydowns had leaked to the junior classes rather than redeeming the triple-A, despite the poor performance of the assets.

Buying off the bottom (hopefully!) is structured credit specialist Magnetar Capital, adding them to a collection which already includes Yate Shopping Centre in Gloucestershire, Perry Barr near Birmingham and Corby in Northamptonshire.

The financing has largely come from Leumi Bank — £18.26m for the Elizabeth portfolio, £38m for Perry Barr and Corby, £30m for Yate, and Magnetar is investing in partnership with specialist asset manager Northdale.

These tickets don’t hold out much prospect of a return to the capital markets for the Elizabeth portfolio (though the Leumi loans are arguably in the sweet spot for a CRE CLO, if it needed the leverage?)

Much discussion, especially in the wake of Covid, has touched on the potential for converting shopping centres to other uses. They often occupy prime city centre locations, attractive for offices or flats, though the footprints make conversion difficult and expensive.

A note from Knight Frank, from May last year, kind of sums up the state of play. It’s titled “Shopping Centres: where did it all go wrong”, saying that rebasing in shopping centre valuations had been brutal, with an aggregate value loss of 80-90% since the global financial crisis.

But there are bright spots, and here’s what probably constitutes the Magnetar trade.

Per Knight Frank, “there are likely to be some sound high-yielding opportunities in our second tier towns and cities where centres now thrive off their rebased rents.”

Retail may have found its floor. Those who wish to shop on Amazon or Temu are already doing that, but plenty of people still appreciate the social experience of shopping. Brands are conscious that retail still gives the best window to consumers, and these centres are still a community asset.

If Magnetar is putting down £10m or so of its own cash for a shopping centre portfolio worth 10 times that only seven years ago, there’s some wiggle room for asset management and value improvement. Time to go shopping!

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