The AI Bubble Just Sprung a Leak — and Kiwi Investors Are Exposed
The AI stock bubble showed its first real crack on June 23, 2026. The Nasdaq opened 2% lower, the Dow and S&P 500 followed, and overnight the contagion spread to Asia: South Korea’s benchmark closed down 10%, Japan’s Nikkei 225 dropped 3.5%. Seven tech companies now make up 30% of the S&P 500’s value — a concentration risk that turns the entire index into a proxy bet on AI. When one of those seven sneezes, everyone’s retirement fund catches a cold.
THE BOTTOM LINE
The June 23 sell-off is the first real stress test of the AI investment thesis. The technology is genuine. The valuations were not. For Kiwi investors, the lesson is concentration risk: if your “diversified” KiwiSaver fund is actually a leveraged bet on seven US tech companies, it is time to find out what you actually own.
What Triggered the Sell-Off
The catalyst was multifaceted. Alphabet dropped 5% on Monday after two high-profile AI researchers left the company, signaling internal instability at the very heart of the AI boom. SpaceX, which debuted on June 12 to massive fanfare, plummeted 16% as it announced plans to raise $20 billion in bonds — despite already pulling in $85 billion through its IPO. The bond offering spooked investors who saw it as evidence that even the most hyped AI-adjacent company cannot fund its infrastructure through revenue alone.
That SpaceX offering lands at an awkward moment. Index providers have already been pushing back on adding SpaceX and AI unicorns to major benchmarks, and the sell-off will not help their case. Meanwhile, the IPO race between OpenAI and Anthropic is still heating up — and Anthropic’s IPO filing now looks like it will hit a far less hospitable market than the one its bankers assumed.
The Federal Reserve’s signal last week that it may raise interest rates to combat inflation added the macroeconomic pressure. When borrowing gets more expensive, the debt-fueled AI infrastructure buildout — Morgan Stanley estimates AI-related borrowing will surpass $500 billion this year, according to reporting by The Guardian — and that debt becomes a liability, not an asset.
Ipek Ozkardeskaya, senior analyst at Swissquote, put it bluntly: SpaceX jumping on the bond train “revives earlier concerns that Big Tech may be spending too much on AI infrastructure and increasingly financing that spending through debt.”
The Dot-Com Echo
The parallels to the early 2000s dot-com bubble are uncomfortable. Then, as now, a revolutionary technology drove valuations to unprecedented heights. Then, as now, a few companies dominated the index. Then, as now, spending was funded by debt and speculation rather than revenue. The difference this time is scale: $500 billion in AI-related borrowing in a single year dwarfs anything the dot-com era produced.
All three major US indices hit record highs this year — Nasdaq up 10% YTD, the Dow up 6% (breaching 51,000 points), and the S&P 500 up 7.3%. Those gains are now looking fragile. The concentration risk is the accelerant: when seven companies make up 30% of the S&P 500, a correction in any one of them drags the entire index down. There is nowhere to hide inside the index itself.
NZ Angle
New Zealand investors are more exposed than they might think. KiwiSaver growth funds and diversified portfolios typically hold 15–30% in international equities, with heavy weightings in US mega-cap tech. When the Nasdaq drops 2% at the opening bell, KiwiSaver balances feel it — not immediately, but in the next unit price update.
The concentration risk flows through to NZ investors in a specific way: most passive international funds track indices like the S&P 500 or MSCI World, both of which are now heavily overweight the same seven AI-exposed tech companies. A Kiwi investor who thinks they are “diversified” across global markets may in fact be making a leveraged bet on the same handful of stocks.
For NZ tech companies, the sell-off is a double-edged sword. Lower valuations for US AI companies could reduce competitive pressure — it is harder for a US giant to justify aggressive expansion when its stock price is falling. But it also tightens the funding environment globally, making it harder for Kiwi startups to raise capital.
The Other Side
This is not necessarily the end of the AI boom. The underlying technology is real, and productivity gains are materializing. What is being repriced is the gap between what AI companies have promised and what they can deliver in the near term. A correction is not a crash — it is the market recalibrating expectations.
The companies that survive will be stronger. The ones that were never more than AI-washing their way to high valuations will not. That is how markets are supposed to work.
FAQ
Is this a full market crash? No. A 2% Nasdaq drop is a correction, not a crash. The 10% drop in South Korea is more serious, driven by the country’s heavy exposure to semiconductor manufacturing — SK Hynix and Samsung both fell 12%.
Should I sell my tech stocks? Panic selling is rarely profitable. The better question is whether your portfolio is over-concentrated in AI-exposed tech stocks. If your KiwiSaver growth fund is 25% in US equities, a meaningful portion of that is in the same seven companies driving the sell-off.
What does $500 billion in AI borrowing mean? Morgan Stanley’s estimate means AI companies are funding their infrastructure buildout — data centers, GPUs, energy — through debt rather than revenue. If interest rates rise, servicing that debt becomes harder, which is exactly what the market is pricing in.
How does this compare to the dot-com crash? The technology is more mature — AI is already generating real productivity gains. But the concentration risk is worse. Seven companies making up 30% of the S&P 500 is unprecedented.
Will the SpaceX and OpenAI/Anthropic IPOs still go ahead? SpaceX is already public and feeling the heat. For OpenAI and Anthropic, a market correction could delay or downsize their offerings. Anthropic’s S-1 is filed, but pricing in a falling market is a very different proposition than pricing in a rising one.
THE BOTTOM LINE
The AI investment thesis just got its first real stress test, and it did not pass cleanly. The technology is real. The valuations were not. For Kiwi investors, the wake-up call is concentration risk: if your diversified fund is actually a leveraged bet on seven US tech companies, June 23 is the day to check what you actually own — before the next 2% drop becomes 10%.
