Do SCF androids dream of future receivables?

Feature-image

Collapsed lender Greensill used AI to identify non-existent receivables which it could borrow against. The concept turned out to be indistinguishable from fraud.

by Nicholas Dunbar

Published: 25 May 2021

“If you are a pharmacy, like any small business, one of your biggest concerns is working capital and cashflow, and of course if you do not have a good working capital solution, you have to go to the bank and borrow money. That is going to be more expensive than the cash less the discount that you are getting from the supply chain finance provider”.

These words were not spoken at a EuroFinance event, but a UK parliamentary enquiry on 13 May. The speaker was former Prime Minister David Cameron, and the ‘supply chain finance provider’ he referred to was Greensill Capital, the lender that collapsed in March and is now facing a criminal probe from UK authorities. Cameron worked for Greensill and was summoned to the enquiry after evidence emerged of his frantic lobbying efforts on behalf of his stricken employer. In the words of one MP, “your reputation is now in tatters, Mr Cameron.”

For years, SCF has been part of the daily lives of the EuroFinance community, from corporate treasurers, to bankers and software vendors. Cameron’s logic has been echoed hundreds of times in presentations and pitches for the product. Even before Greensill’s collapse, SCF was facing criticism for lack of transparency and for legitimising slow payment of smaller suppliers by multinationals.

Yet Greensill has undercut the most potent growth message of SCF: that some kind of trillion dollar fintech miracle is on the horizon. Cameron expressed this premise in his testimony as follows:

“What Greensill were doing with partners such as Oracle or Taulia, or Textura in the past, was using the information in a company’s ERP to make sure you could extend credit to suppliers, and indeed to employees, faster, and I think that does qualify as fintech, because the fin is the access to the deep capital markets and the tech is using the ERP to make credit decisions better and faster”.

Greensill founder and CEO Lex Greensill explained how he piggybacked on Taulia’s access to corporate SAP and Oracle ERP systems in his own testimony to the parliamentary committee on 11 May. “We were on the fourth generation of our technology at the time that our company went into administration, which was substantially reliant on artificial intelligence and machine learning to deliver the solutions that we did”, Greensill explained. “We were using data inside corporate systems on customers or suppliers to unlock capital.”

So confident was Greensill in his AI that he used it to securitise receivables that didn’t yet exist – so-called future or prospective receivables. To the frustrated British MPs questioning him, this amounted to pretending that unsecured loans were secured, prompting allegations of fraud.

Even if we accept Greensill’s denials, the damage was done and here consider the other half of Cameron’s formula – the ‘fin’ side of fintech. For Greensill the ‘deep capital markets’ that Cameron spoke about turned out to be a Greensill-owned German bank and a Credit Suisse fund. Lacking hard evidence of receivables, Greensill was forced to rely heavily on external insurance providers in a way that Lex Greensill himself acknowledged was a mistake.

Now, amid the Greensill fallout, the industry is retrenching. The idea of applying SCF to public sector procurement is discredited: after all, a government that issues its own currency doesn’t need working capital. It should simply pay small suppliers promptly rather than enriching intermediaries like Greensill. And those who were once close to Greensill, such as SCF vendor Taulia, have pivoted swiftly away from it towards traditional bank funding channels.

With sufficient transparency, SCF will continue to have a role in balancing the working capital needs of multinational purchasers and their suppliers. But for now, trillion dollar AI dreams are out of the running.