Receivables Finance: The Verification Gap After First Brands

By Igor Zaks, CFA – President, Tenzor Ltd.

In the 2022 World Supply Chain Finance Report, I wrote that the Greensill collapse had “served as a litmus test for policies and controls needed to operate supply chain finance and receivables businesses.” I argued that the industry needed to look behind the product’s name – SCF, credit insurance or securitization – and understand its structure, underlying assets and risk-management procedures.

Five years after that collapse, the test is being administered again. The results are not encouraging.

Over the past ten months, a single fund – Jefferies’ Point Bonita Capital – has been at the center of three separate situations in which the documents supporting receivables financing have come into question. The emerging pattern cannot be explained solely by bad luck or conventional credit selection. It points to a structural weakness in verification, servicing and control that also featured prominently in Greensill.

What Greensill actually taught us

The convenient story about Greensill is that it was a story about one charismatic financier and one over-leveraged steel group. It was considerably more than that: extreme concentration to GFG Alliance, insurance non-renewal, liquidity and funding dependence, regulatory intervention, future-receivables financing, and governance failures that FINMA later found extended into Credit Suisse’s own risk management and organizational structures.

But one strand deserves more attention than it received, because it keeps recurring.

Credit Suisse extended a $140 million emergency loan to Greensill in October 2020, partly supported by invoices Liberty Commodities had sold to Greensill. Several businesses named on those invoices subsequently told the Financial Times they had done no business with Liberty. Bloomberg had earlier reported that four banks stopped working with Gupta’s commodity-trading operation beginning in 2016 over concerns about bills of lading and other paperwork.

Read that sequence carefully. Some financing institutions identified documentary concerns years before the collapse and exited. Others continued providing funding. The information asymmetry was not simply between the market and the borrower. It was also between the institutions that independently checked and those that did not.

The UK Serious Fraud Office opened its investigation into GFG’s financing arrangements with Greensill in May 2021, covering suspected fraud, fraudulent trading and money laundering. It remains open. GFG has consistently rejected any wrongdoing.

I made a related point about servicing. Many receivables securitizations, particularly where continuity of collections is material, incorporate a backup-servicing arrangement under which a qualified third party can assume the servicer’s functions. In my 2021 and 2022 analyses, I noted that no effective pre-positioned arrangement appeared to have been in place at Greensill. Investors were left depending on the originator to service the very assets securing their investment. That dependency question runs through everything that has followed.

First Brands: from theory to documented finding

First Brands Group filed for Chapter 11 in September 2025. Jefferies disclosed that Point Bonita held approximately $715 million in First Brands-related receivables within an approximately $3 billion trade-finance portfolio. First Brands acted as servicer and stopped directing timely transfers on 15 September 2025.

The court-appointed examiner’s report moved much of this from inference to record. According to that report, North American factoring programs were “less transparent” than European ones, relying on spreadsheets produced manually and uploaded to a supply chain finance platform. Lenders “typically did not see the actual underlying invoices or verify the data against the invoices stored in [First Brands’] database.” Specifically, Leucadia “relied on invoice-level data presented in PrimeRevenue” when purchasing receivables and “did not independently receive or review underlying” documents.

There is no suggestion of wrongdoing by PrimeRevenue, which has said it provides technology platforms for financing programs, not credit underwriting or invoice verification, that it was not the servicer in the programs referenced, and that the examiner did not contact it.

That clarification matters, and it is precisely the point. The platform did what platforms do. The question is what the funder did on top of it.

The geographic contrast in the examiner’s findings is striking, although it should not be overstated. According to First Brands executives interviewed by the examiner, European teams at Citi and BBVA rejected proposed arrangements after identifying information concerns. At the same time, the North American programs proceeded under materially different structures. The public record does not establish that the institutions reviewed identical documents or were offered identical facilities.

Even with that caveat, it resembles an important feature of the Liberty Commodities history: some institutions reportedly identified documentary warning signs and withdrew, while other financing continued.

Radiant World and Sapphire Minmetals: the pattern accelerates

The current situation is at a much earlier and genuinely disputed stage, and it should be described with care.

Bloomberg reported that Vitol and Cargill had stopped trading with Radiant World, an iron ore trader, and that Glencore had ceased entering new business, amid concerns that Radiant had provided banks with falsified documents relating to iron-ore trades. Reuters separately confirmed the trading pullback and Intesa Sanpaolo’s approximately EUR 200 million exposure, against which Intesa said it had booked provisions and that the position was largely covered. Still, Reuters did not independently verify the document allegations. Radiant World has denied the allegations, stating the claims are “inaccurate and unsubstantiated” and that its business continues to operate normally. Point Bonita’s exposure is reported at less than $300 million.

Bloomberg has also reported that Rabobank stopped financing Radiant in early 2020 after an internal investigation identified multiple trades involving falsified bills of lading, and that ING stopped financing the company around the same period over similar concerns. Radiant has denied allegations of wrongdoing.

On 4 August 2026, Bloomberg reported a further development involving a second iron ore trader, Sapphire Minmetals. Bloomberg reports that Jefferies recently contacted Vitol, a named counterparty on Sapphire-related invoices, and was told that some of those invoices were not genuine. Sapphire’s chairman “categorically and strongly” denied the report, calling it “untrue and unsubstantiated.” Jefferies and Vitol declined to comment, and Reuters said it could not independently verify Bloomberg’s report.

Nothing here is a proven finding of fraud, and I want to be explicit about that.

But the reporting raises a question the industry cannot avoid. The public record does not establish whether equivalent confirmation or other independent validation had previously occurred before the relevant advances. That is now one of the most important unanswered questions in this case.

If decisive independent verification occurs only after a portfolio enters wind-down, it is no longer operating as a preventive control. Whether that was the case here has not yet been established publicly. But it is the question every allocator and risk committee should be asking about their own books this week.

Direct confirmation with the named debtor is one of the clearest verification controls in receivables finance. It need not be applied identically to every transaction – non-notification factoring and portfolio facilities may instead rely on ERP integration, sampling, debtor audits, electronic acknowledgments, delivery evidence, controlled accounts and borrowing-base testing. But where direct confirmation is absent, the funder needs equivalent independent controls appropriate to the structure, risk and concentration. What it cannot do is treat documents supplied by the financed party as self-verifying.

What actually connects these cases

These are not mechanically identical transactions. First Brands involved automotive receivables with the borrower acting as servicer, alongside genuine credit and liquidity deterioration. Radiant and Sapphire involve commodity trade finance and remain unresolved. Greensill involved approved payables, future receivables, insurance non-renewal and related-party exposures across an industrial group.

What connects them is not that no institution ever checked. In both Liberty Commodities and Radiant World, reporting indicates that some institutions identified warning signs and exited years earlier. The more troubling commonality is that, for the institutions that remained exposed, weaknesses in transaction validity, receivable ownership, servicing or cash control were not prevented – or did not become decisive – until payment stress, regulatory intervention or insolvency.

The relevant verification questions are consistent across the cases: did the underlying transaction occur; was the invoice genuine and enforceable; was the receivable owned by the seller and free of competing assignments; and were the resulting collections directed into an adequately controlled account?

Four process weaknesses recur:

Concentration presented at the wrong level. Point Bonita’s investor materials reportedly emphasized exposure to numerous named obligors, while First Brands ultimately represented roughly a quarter of the trade-finance portfolio. In receivables finance, obligor diversification can conceal concentration to a single seller, originator or servicer unless each dimension is measured separately. A portfolio of invoices naming Walmart, AutoZone and NAPA is not diversified if a single seller generated all of them.

Documents accepted from the financed party. In 2022, I wrote that “traditional” SCF converts performance risk into essentially buyer credit risk only through unconditional buyer confirmation that the invoice is valid and payment approved. Absent that confirmation or equivalent controls, seller-performance and invoice-validity risk remains. Standard trade-credit insurance generally covers non-payment of a valid, compliant debt; it does not ordinarily eliminate those underlying risks. An invoice presented by the seller and unconfirmed by the debtor is not blue-chip credit risk wearing a blue-chip name.

Technology mistaken for control. I warned in 2022 against confusing the use of technology with an AI-driven risk process, and asked what data any model was calibrated on. The current version of that error is subtler: a platform provides visibility into data the financed party uploaded. That is not verification. Platforms are rails. Verification is a decision the funder must make and evidence.

Servicing and cash control are left with the counterparty. Where the financed party also controls collections, the funder depends on the entity it is exposed to for the information proving that exposure is sound. Greensill appears not to have had an effective pre-positioned backup-servicing arrangement. First Brands acted as its own servicer, on Jefferies’ own disclosure. This dependency is not an operational footnote; it is a mechanism by which problems can stay invisible.

The uncomfortable question for allocators

Point Bonita’s investor materials reportedly highlighted a record of no down months.

A record of no down months in receivables finance is only as strong as the controls beneath it. Unverified paper does not generate volatility – it generates apparent stability, right up until it does not. Smoothness is not evidence of quality; in this asset class it can be evidence that little is being tested.

Investors, allocators and boards should be asking a different set of questions than “what is the credit quality of the obligors”:

• Who supplied the documents supporting each receivable, and who independently validated them?
• What independent validation was performed before funding, how frequently was it applied, and can the results be evidenced at the appropriate transaction or portfolio level?
• What prevents the same receivable being assigned twice, across facilities and across funders?
• Where does the cash land, and who controls that account?
• What is the concentration by seller, originator and servicer, not just by obligor?
• If the originator failed tomorrow, who services the book – and has that arrangement been pre-positioned and tested?
• Which counterparties or financiers have already exited this name, and why?

That last question is among the cheapest diligence available, and in both Liberty Commodities and Radiant World it was answerable years in advance.

This is a solvable problem.

None of this is an argument against receivables and supply chain finance. Well-structured receivables transactions remain among the more resilient assets in credit – I said so in 2022, and I believe it more strongly now. Trade receivables securitizations came through 2008 with a good record precisely because the discipline was real: genuinely diversified pools, stress-tested over-collateralization, meaningful equity ahead of investors, and rating agencies that analyzed the portfolio before crediting external support.

These episodes do not establish that receivables finance is inherently defective. They do, however, repeatedly expose process failures in verification, concentration measurement, servicing and cash control, alongside the distinct credit, liquidity, governance and legal problems present in each case. The process failures are the fixable part – and fixable in advance, rather than in a wind-down.

That requires specialized operational and structural expertise that is unevenly distributed across fund managers, banks and investment teams. The traditional asset management culture was shaped by portfolios of public securities: light-touch, top-down, comfortable with summary reporting. Receivables finance demands the opposite: review at the level of individual exposures, operational process, legal structure, insurance terms, and originator diligence.

That is exactly the work Tenzor does.

We conduct operational due diligence on receivables and supply chain finance programs, portfolios, and originators. We review verification and control frameworks against the failure modes that have actually caused losses, not theoretical ones. We assess concentration on the dimensions that matter, stress-test servicing and cash-control arrangements, and evaluate whether insurance and structural protections would perform as intended when they are needed. We work with investors before they commit capital, with funders reassessing existing books, and with platforms and originators who want their controls to withstand scrutiny.

I have spent my career working across banking, structured finance, corporate working-capital programs and receivables finance, including roles at Citigroup, Eurohypo/Commerzbank, Daiwa and Dell; building and leading receivables-finance platforms at SCF Capital (2006-2009) and, more recently, 40Seas; managing a distressed portfolio through a crisis; and advising funds and originators through Tenzor. I wrote about Greensill’s control failures in 2021 and revisited them in the 2022 World Supply Chain Finance Report. I have covered First Brands and the other recent cases extensively in the World Factoring Yearbook. Both Bloomberg and the Wall Street Journal have quoted my analysis of these situations. I would rather help institutions avoid becoming the subject of the next postmortem.

If you are allocating to, managing, or originating receivables and supply chain finance – and you cannot currently evidence what independent validation was performed before capital moved – that is the conversation worth having now. At the same time, it is still diligence rather than litigation.

Order the World Factoring Yearbook 2026 here.

Igor Zaks, CFA, President, Tenzor Ltd. tenzor.ca | Toronto

This article discusses matters that are the subject of ongoing litigation, regulatory investigation and press reporting. Allegations described are unproven and, where noted, expressly denied by the parties concerned. Reporting attributed to Bloomberg, Reuters and the Financial Times is described as reported. Nothing here constitutes legal, investment or accounting advice.