Blain’s Morning Porridge August 17th, 2026 – AI Driven Markets and Liquidity May Part Company!

“Euphoria one day, the sound of banks crashing the next…”

The factor that’s enabled the extraordinary success and longevity of the AI bubble has been the willingness of the markets to finance it. What happens if the liquidity machine driving AI were to suddenly stop? As the limits of bond markets, private credit, and using insurance companies to park risk, become increasingly apparent – what would a liquidity event do to current markets? Ouch!

LINK TO PODCAST

Key Takeaways:

  • Equities and bonds are diverging. Equities think the party continues. Bonds believe inflation and higher rates are nailed on. They can’t both be right
  • Liquidity if the key factor driving markets – not belief in AI. If it dries up, then the whole market wobbles.
  • A Liquidity Event is probably hiding in plain sight – perhaps a US treasury auction that goes awry, or an insurance company taking a thumping on a credit loss.
  • Risk has not disappeared – it has been transferred from the banking sector to the asset management sector.
  • Because investors think Elon Musk will make a digital data centre work doesn’t make it a fact – until, maybe, he does.

Markets are neither smart nor clever. They represent the weighted sum of the mood, sentiment, outlook and experience of every participant, but they can’t actually tell the future – they merely predict/reveal what the crowd feels it most likely to be. That’s the theory… In practice, markets comprise folk who make mistakes, get sucked into narratives, and often aren’t aware of the distortions that play upon them in markets.

Market participants tend to follow the flow (the trend is your friend) because following the majority decision is an easy call to make. (It explains why it’s so hard for contrarians to be proved right!)

However, this morning, global markets feel conflicted. The stock markets are off to the races, still climbing on what is perceived to be a rising corporate outlook and the potential upside for stock markets from AI – even though it’s increasingly difficult to see where the returns will be generated from given increased competition and the scale of spending in the sector. In contrast, the bond markets are distinctly gloomy about inflation and rates. Bond markets are seldom mistaken for a ray of sunshine.

Such is the success and power of the AI narrative that no one is looking much beyond it.

The place markets should be looking is liquidity. The liquidity that’s fuelled the AI boom has been phenomenal. From the massive infrastructure borrowings of the hyperscalers, the Space X IPO, to the $500 bln Chips purchase facility for Nvidia, US lawmakers are taking pride in the strength of US capital markets. It’s been capital markets liquidity that’s fuelled US bond and stock markets for the last three decades, building the USA’s stock market exceptionalism.

While easy money makes growth capital available, it’s only one part of the equation alongside invention and innovation. And it’s also true that if money is too cheap, then it will be misused. But give the markets credit: it’s been the ability of the US capital markets to provide liquidity to fuel US markets that has been the foundation of the economy’s strength. Historically it has also been a vulnerability – if fuels speculation.

What would happen to markets if the supply of capital that’s been pumping US markets ever higher was to suddenly stop?

While stock markets seem determined to rise, bond markets are looking for reasons to hide. Whether it’s the threat of higher interest rates as the oil dislocations from Hormuz roll ever on, or the possibility of further supply chain ructions as a result of tariffs, bond markets are turning increasingly bearish. US Treasury bond yields are at decade highs. Junk bonds have been edging higher. Buyers are concerned about buyers – who is going to keep absorbing the supply? From Governments and from Hyperscalers?

This morning the FT is warning about further strains on Private Credit Portfolios – impairments (that great expression from banking) are rising! If defaults are rising, then lending is likely to become… tougher.

As I described last week in Complexity in AI Financing – you can’t just make risks inherent in any capital markets disappear – though you can transfer and transform them. If something wicked this way comes for global markets, I reckon it’s going to be a liquidity event that triggers a market shakedown. And when it happens, we’ll all be shocked how easily it’s been hiding in plain sight.

Markets are supposed to be about common sense – reaching agreement on prices based on the weighted expectations of buyers and sellers. But we know markets are also vulnerable to distortions. It was easy money that drove the big distortion of the QE era – effectively creating inflation in bond and stock market prices. Easy money simply fuelled speculation which is why the 2010s and 2020s have been so susceptible to speculative investment themes – from Bitcoin, NFTs, the metaverse, and whatever the next fad in tech will be.

Too many times, I’ve written about how equity players tend to leap into these narratives – the next new, new thing which oft prove to be short-lived bubbles. Stock buyers like nothing better than a good story, but when you see adverts for 3x Leveraged Space X ETFs in the mainstream press – you just kind of know this is going to end badly.

When I look at the market and see the volumes being pumped into new deals, or at levered retail investments, it feels like excess liquidity is a major factor driving markets – anything and everything remotely AI linked is fundable.

The market gets it. Open AI and Anthropic (purveyors of the finest artificial intelligence on the planet), are rushing through their IPOs, determined to get market as soon as possible – before it pops. I suggested the Space X IPO was a signal for the top – but although it’s been volatile, and underperformed the S&P500, it’s not yet burst completely. (Although the $135 IPO price to the $225 top and the $105 bottom has been quite a journey!)

There are investors who genuinely believe the billions being spent on AI will generate trillions in returns… but the smart money is getting nervous. Abundant liquidity, easy leverage and capital markets willing to fund anything improves the mood no end. From the 1920s, 1980s and 2007, its always the same… massive riches will arrive tomorrow as long as abundant liquidity and leverage is available to fund it all… Where are you going to sit when the music stops?

Back in the rational and calmer world of bonds… there are reasons to be fearful. The signals I’m watching are grey and bleak: rising government bond yields, rising junk yields, the buzz around the AI infrastructure bubble “crowding out” other issuers, and now a slew of stories of how Private Credit players have been loading captive insurance companies with illiquid assets! (Again in the FT – it would appear some US private capital firms have been loading up their own insurance vehicles with dodgy deals! Gosh… who would have thought…)

Way back in 2010 I was warning the next crisis would not be another series of bank runs. The regulatory authorities were swift to crack down on banking risk. Today the Central Bankers run bank stress tests, which sometimes seem more concerned to show us how stable the banking system, rather than test the whole financial system.

The great rule of markets is you can’t eliminate risk, but you can transfer or transform it. From 2008 onwards banks were regulated to transfer risk from the banking sector to the investment/asset sector. Insurance companies are like mirror images of banks – banks die quickly because of liquidity crises meaning they can’t finance their assets. Insurance companies die slow when they can’t sell assets to meet liabilities. And that may be the form the next crisis takes – a sudden realisation that credit and market losses are sitting offshore in supposedly solid investment and insurance funds.

What’s been known for years is how Private Credit firms have being buying up or creating insurance firms who’ve stopped buying liquid government bonds to match their long-term liabilities and now buy complex private credit assets instead…. Good luck trying to liquidate these portfolios when the next crisis comes a ‘knocking!

The thing about markets is they are supposed to be smart because are the weighted sum of the opinions of every participant. The mythos behind efficient markets is that the larger the number of buyers and sellers, the more perfect the weighted sum of all these opinions will be. The same thinking drives the probability markets – if enough people expect something to happen… the thinking is it probably will. But events and consequences are usually discrete.

There are expectations/result correlations in some circumstances – like voting. People will bet on how elections play out based on the continuous available information from polls, and now probability, markets. To a degree the probability market becomes self-reinforcing where the information is available.

However, it should not work in markets. The likelihood that SpaceX will perfect the “Space Based AI Data Centres” upon which the current lofty expectations of its future value are based, depends upon on a whole series of future discrete events occurring – which will result from the interplay between engineers, physics and Elon Musk threatening to fire staffers who don’t deliver. The fact the market believes Musk will deliver has not one iota of influence on what will actually happen.

Bill Blain

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

Meanwhile, don’t forget about my new book: The Battle for Hamble

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