Blain’s Morning Porridge August 13^th, 2026: Complexity in AI financing will not make the risks disappear.

“Complexity does not eliminate risk – it simply hides it in plain sight.”

The AI infrastructure build out already dominates credit market funding. Now Nvidia will effectively securitise its chips to its market though a panel of private capital markets firms that lend it credibility. The worry is complex and appealing deal structures will simply hide the risks inherent in AI – and increase the likelihood of a correlated crash when something inevitably breaks.

LINK TO PODCAST

Key takeaways

• Financial complexity will make AI easier to finance – but it does not change the economic risks.
• The securitisation of Nvidia chips has the potential to accelerate the AI boom – but comes with increased correlation risks.
• Nvidia is “guilty” of circular financing – funding its customers to buy its chips and using the markets to absorb the risks.
• There are questions about chip longevity – how soon will they become obsolete or be replaced by competitors?
• The degree to which Nvidia risks now links hyperscalers, datacentres, energy, water, and the AI industry is enormous, and could destabilise the economy in a crisis. There are multiple points of potential failure.

What’s not to like about the new Nvidia $500 bln Chip Funding Programme? It’s going to launch the AI revolution stratospheric! The upside is AI firms won’t have to front up billions to buy the chips they will need, they can borrow against them instead. They won’t have to build infrastructure or data-centres – they can rent capacity from the hyperscalers. Therefore… so the thinking goes … the securitisation of chips will trigger an explosion of AI firms able to enter the market – ensuring the viability of the infrastructure and data centre build out.

Simples. Joy unconfined. All problems in AI capital solved.

Alchemy – the dark art of turning base metal into gold – has always been the core dream of the financial world. To achieve returns, clever financial wonks are cooking up complexity to create structures to finance and enable the AI infrastructure build out. I’ve been told to stop thinking of Nvidia chips as mere GPUs – think of them as a new financial currency. Nvidia CEO Jason Huang even called his chips “fungible”. Gosh. How clever.

Except… of course, they aren’t! (Clue: “Danger, Danger, Will Robinson, Danger” has been blaring in my head since I read about the programme earlier this week.)

What can possibly go wrong? Gosh… where do I start…?

There is one immutable law in finance: you can transform, transfer, slice and dice, and hedge against risk, but you can never make it go away completely. Risk is risk. In any transaction, no matter how large, simple, complex or small, there is always risk. You can hide risk behind complexity and structure… but, usually it is hidden in plain sight. Risk never goes away. It is always there. The trick is to identify it – and then mitigate it.

The second unbreakable law is that everything is connected. When you make Nvidia chips the centre of the AI ecosystem, it simply means a failure at any point: in energy, in datacentres or even water for cooling, now becomes a failure point for that whole system. Complexity results in increasingly correlated risks!

I have been here before. I remember sitting with my sales team at a leading bank 19 years ago in early 2007 explaining why the new CLO cubed we were launching was such fantastic value… and dismissing their concerns entirely. I have learnt much since then…

Nvidia has partnered up with the New Giants of the Private Capital Markets – the private credit firms: Apollo, Blackrock, Blackstone, Brookfield and KKR, plus investment bank Goldman, to create the new $500 bln financing platform. It will enable third party investors to lend to the AI market. AI firms will be encouraged to borrow against the value of the chips they order from Nvidia. It’s securitisation, vendor and trade finance rolled into one. It closely links capital and production. Perhaps too closely.

The six private capital firms won’t be putting up only their own money onto the platform. Even as I write, their salesmen will be selling the deal to the securitisation/tech desks of the world’s leading Sovereign Wealth Funds, Pension Funds and other asset managers – persuading them this is the perfect entry to the ultimate rally. They will be promising stellar returns and making much of the security of the structure – including Nvidia taking the first loss trance of 25%!

I’ve been involved in the securitisation markets almost since the start of my career. As a young journalist at Euromoney in 1987 I interviewed the guys behind the first Euromarket mortgage securitisations. When I resumed my career as a banker I put together Auto and aviation backed deals. I was closely involved with the financial alchemy behind the collateralised debt and loan markets – which caused such angst through the global financial crisis in 2008. Most deals do wonderfully well – but some exploded.

I first heard about chip receivables a few years ago – US private credit funds lending tech start up’s the capital required to finance their compute purchases – the trick was to secure lending by using the chips as collateral. Given the speed at which chips have evolved – effectively built in obsolescence, and the lack of actual profits (many tech firms view customer acquisition as their critical metric) – I wondered how the firms would actually repay the loans, and what value loans “secured” on second hand chips would actually have?

Most banks won’t lend against chips. They depreciate too fast. In terms of mitigating the obsolescence of chips – Nvidia has said will guarantee the first 25% loss of residual value in structures linked to its chips. That’s a massive positive for investors.

For the funders, the AI firms buying Nvidia chips, the key issue will be how they pay the debt on the chips? What happens when their profits fail to meet the expectations, and they default? Or if business conditions go against them in terms of competition, or the energy costs in datacentres, or new ways of doing AI with other chips emerge?

What happens if the actual demand for AI underwhelms? If American Closed-Weight LLMS fail to garner the market share expected of them? If firms decide the best AI for them is an open-weight system, based on an office Mac Studio? Or perish the thought… buy Chinese instead? What if China floods the market with cheaper compute?

What happens if Nvidia is supplanted as the best Chip maker? Unlikely say the shills – because Nvidia will be able to sell more chips, their chips will be getting better faster than other firms, thus it will remain the industry leader. (Tell that to the Wright Brothers who saw their monopoly in powered flight disappear in 10 years!)

The bottom line is the whole Nvidia deal is predicated on Nvidia continuing to be Nvidia – and that nothing else makes them obsolete. That is the risk – that Nvidia does have the longevity the new platform assumes. When some 15% of total corporate debt is now linked to AI – are we close to a stage where it all goes horribly wrong at speed if something breaks?

Out of time and back to the day job…

Bill Blain

Author of the Morning Porridge
CEO Windshift Capital
Advisor – Spitfire Strategic Capital

Meanwhile, don’t forget about my new book:

You can read a review on the Society of Professional Economist’s website here.

The Battle for Hamble is a proper grown-up examination of how economies fail: a tale of Greedy Corporates, Bad Planning and Economic Illiteracy. It uses a wholly unnecessary Gravel Quarry in the middle of a prosperous village to illustrate the multiple failings and abdications of responsibility that have created Broken Britain. It’s about bureaucracy, money, exploitation and shareholders vs stakeholders. It’s about social injustice – asking why it’s ok to put 6000 jobs at risk so corporate bosses can reap bigger bonuses! Whether is the Hamble Quarry, HS2 or the abysmal state of UK armed forces due to bad procurement, read the book to understand how its broken process and bureaucratic indifference that is sinking Britain.