That was quick. Half the year gone already but no World War III, no AI ending humanity and no gains for all those crypto lemmings who increased the wealth of the Trump family by $1.4 billion. The Donald deftly sidestepped the crypto shake down with the reassuring deflection of a practiced mobster – “The stock market is going up…Everybody’s profiting”. Sure, Jan. Between Love Island and the upcoming weekend sports-fest one can understand people lacking a little financial focus. So, I will keep it brief today. I’d like to take a look at a number of technology sector financial milestones which have been achieved and then flag a couple of unintended consequences, and probably pension risks. First, the milestones….
- Tech-heavy Nasdaq Index gained 20% in H1 vs S&P 500 up 9.5%.
- Semiconductor/chip sector went rocketed 82% in the same 6 months (Nvidia, Broadcom, Intel etc)
- Memory chip stocks like Sandisk, Micron, Hynix and Samsung are up a whopping 120% in H1.
- Research house, Gartner, say AI spending will hit $2.6 TRILLION in 2026.
- The AI hyper-scalers – Google, MSFT, Amazon and Meta – are set to spend $650 billion on AI infrastructure in 2026.
- The combined weight of AI-focused stocks across hyperscaling, semiconductor chips, power, hardware and software tots up to 51% of the total value of the S&P 500 index.
- Nine major AI companies accounted for almost half of global technology borrowing, raising $122 billion in corporate bonds in a single year to fund data centres and infrastructure.
So, my first observation based on these milestones is that, if your pension is tracking global/US stock markets, then there is a strong possibility you are ‘running’ a significant bet on AI without actually realising it. It’s what the pensions/wealth industry might refer to as a ‘concentration risk’. And, I think the following headlines are flagging a few other AI risks right now….
- OpenAI Leans Toward waiting Until Next Year For IPO – New York Times
- Tesla Caps Employee AI Spend At $200 Per Week After Adoption Push – The Information
- OpenAI in early talks to give 5% stake to US government – The Guardian
OpenAI, as a reminder, is attached to almost $1 trillion of AI infrastructure projects and the ‘mood music’ in the above headlines is not great. These projects have been funded by trillions of equity and debt from technology and banking partners. So, these partners must be wondering why OpenAI feels the need to grease Donald Trump’s tiny toddler fingers. I’m wondering too, but speculation gets us nowhere. Of course, the complete anti-Donald antidote is truth, numbers, facts and genuine science. So, I was intrigued to come across some excellent research by former colleagues of mine at Quant Insight. These guys use big AI computational power and principal component analysis (PCA) to strip out all the ‘noise’ attached to the pricing/trading behaviour of financial instruments in the equity, debt and FX markets. The benefit of this huge analytical undertaking is to identify the key factors/drivers of a share price or bond price in the current market environment/regime. This is what they found was driving the $10 trillion semiconductor sector ETF (SOXX) which rocketed 80% in Q2 alone….
It turns out that the biggest external (macro) factor driving the share prices of semiconductor companies was….. lower cost of corporate borrowing. Now think about these companies involved in heavy capex manufacturing and infrastructure activities. A glance at the financial milestones above and trillions of dollars of planned investment spend means these tech companies need external funding given their own revenues and cash flow can’t keep up with the pace of investment required. This means technology companies are now borrowing which was never really a feature of these high margin/cash flow companies previously. For pension funds this ALSO means the whole AI infrastructure story is not just a stock market story. Hidden behind the headlines, there is a borrowing, credit, balance sheet story. Now, think about that 51% exposure of the S&P 500 index to AI. You think you’re getting equity and AI exposure but….. you’re also acquiring an exposure to a credit (lending) book as large as many dedicated private credit funds. Now check out the recent headlines on private credit funds.
Actually don’t. Enjoy the weekend sport first!
