<h2>The Paradox: Record Earnings and a Falling Market</h2>
<p>Micron Technology reported on June 24 2026 that its quarterly revenue had quadrupled to 41.46 billion dollars. That is not a typo and it is not a rounding effect. Quadrupled. From 9.3 billion dollars in the same quarter a year ago to 41.46 billion — a rate of growth that has been achieved by almost no large company at any point in the history of modern capitalism. Earnings per share came in at 25.11 dollars adjusted against expectations of 20.49. Data centre revenue hit 25 billion dollars in a single quarter. Management described free cash flow as the highest in the company's history. The stock jumped 15 percent on Thursday. And the Nasdaq still fell for the week. How is that possible?</p>
<p>The answer requires separating two questions that investors frequently conflate: is this a great business, and is the stock fairly valued? Micron at 41.46 billion dollars in quarterly revenue is unambiguously a great business. The question the market was asking this week was whether the stock price — and the prices of dozens of adjacent AI-linked companies — had risen far enough, fast enough, that even extraordinary results like Micron's were already priced in, leaving little room for the further gains that current valuations imply. Rick Gardner of RGA Investments described it as a healthy pullback since many tech stocks have become overstretched. David Makaryan of Alpha Pacific Group said traders are reassessing technology stock valuations while some investors are taking profits following a strong rally. Neither of those characterisations says the AI story is wrong. Both say the price at which AI stocks were trading going into this week reflected expectations that were very high and therefore very sensitive to any signal that reality might fall short of them.</p>
<h2>What Triggered the Specific Selloff This Week</h2>
<p>Several distinct triggers combined this week to produce the technology correction. The first was Bank of America's note on Monday raising concerns about AI chip valuations and warning that earnings expectations for technology stocks heading into July's results season were too high. BofA's analyst teams are among the most widely read on Wall Street, and a note from them expressing valuation concerns in a sector that has risen 27 percent in three months is enough to make institutional investors who have been considering taking profits decide that this week is the week to do it. The second trigger was the news from South Korea that SK Hynix and Samsung — the two largest producers of the AI-grade memory chips that underpin Nvidia's GPU cluster performance — both fell more than 12 percent in a single session, producing a 10 percent decline in South Korea's Kospi index. Korean semiconductor stocks are among the most direct proxies for AI infrastructure demand and their sharp decline sent a signal to global chip investors that reverberated from Seoul to Silicon Valley.</p>
<p>The third and perhaps most emotionally powerful trigger was Apple's MacBook and iPad price hikes on June 25. Apple's announcement that it was raising MacBook prices by up to 300 dollars because of the AI-driven memory shortage was a clear and visible demonstration of the second-order costs of the AI boom — costs that flow from the data centres and their GPU clusters into the supply chains of every consumer electronics company in the world. When Apple raises MacBook prices because Micron's memory is now too expensive for consumer electronics manufacturers to source at acceptable cost, investors in Apple stock sell. Apple fell 6 percent, its worst day since the tariff shock of April 2025. And when Apple falls 6 percent, the Nasdaq falls with it, because Apple is one of the Nasdaq's most heavily weighted components. The chain of causation from AI data centre demand to memory shortage to MacBook price hike to Apple stock decline to Nasdaq weekly loss is not metaphorical. It is a direct mechanical connection.</p>
<h2>Is the AI Investment Thesis Broken? No — But It Is Maturing</h2>
<p>The most important question for investors observing the June 2026 technology correction is whether it represents a fundamental challenge to the AI investment thesis or merely a pause in a bull run that remains structurally intact. The weight of evidence available this week suggests the latter. Micron's 41-billion-dollar quarter is not a sign that AI infrastructure spending is slowing. Quite the opposite — it is the most direct possible confirmation that hyperscalers are spending at a scale and pace that is transforming the economics of the entire memory industry. Nvidia's recent earnings, also extraordinary, told the same story from the GPU side. The capital expenditure plans of Microsoft Amazon Google and Meta — collectively committing to hundreds of billions of dollars in AI infrastructure in 2026 — have not been revised downward. The AI buildout is real, it is large, and it is accelerating.</p>
<p>What is maturing is the market's willingness to price that reality indefinitely forward at ever-higher multiples without asking harder questions about specific companies' earnings trajectories in the near term. The S&P 500 technology sector was up 27 percent in just three months before the correction. The Philadelphia Semiconductor Index was up 64 percent year-to-date before this week. Those are not normal rates of appreciation and the stocks that produced them had priced in a great deal of future earnings growth that now needs to be delivered, quarter after quarter, in actual reported results. Micron delivered. The market's reaction — a 15 percent one-day gain — was positive. But the sector as a whole still fell because the investors who had been considering taking profits for weeks found in the Apple and Microsoft price hike announcements a specific signal that the AI boom's costs were beginning to create headwinds at the consumer level that might eventually affect the demand for the hardware and services that the AI boom is fuelling. Whether those headwinds are modest speed bumps or genuine threats to the AI infrastructure spending trajectory is the question that will determine whether June 2026 looks like a correction within a bull market or the beginning of a larger recalibration. The answer will come from the companies reporting earnings in July, and from the diplomatic situation in Iran, and from the Federal Reserve's next meeting, and from the hundred other variables that make markets simultaneously fascinating and infuriating to anyone trying to understand them.</p>
<h2>What Investors Should Actually Watch</h2>
<p>For investors trying to navigate the post-correction environment in AI and technology stocks, three indicators are worth watching more closely than any individual stock price movement. First, the July earnings results from Microsoft, Amazon, Google, and Meta — the hyperscalers whose capital expenditure decisions are the primary driver of AI hardware demand. If these companies report that they are slowing AI infrastructure spending, or that the returns on their AI investments are lower than expected, that is a genuine fundamental signal that the AI buildout is moderating. If they continue to accelerate spending and report strong AI-related revenue, the correction of June 2026 will look like exactly what Rick Gardner called it: a healthy pullback in an overstretched trade. Second, the trajectory of oil prices and the Federal Reserve's response to the improving inflation picture. Lower oil means lower inflation means more chance of eventual rate cuts means a higher fair value for growth stocks at any given level of earnings. The PCE reading of 2.1 percent on June 27 was the most important single data point of the week for the valuation arithmetic of technology stocks, and it was barely covered in the context of the Apple and Micron stories. Third, the IAEA inspection situation in Iran. If Grossi's implicit 10-day deadline passes without inspection access and the IAEA reports non-compliance to its board, the risk of renewed Hormuz disruption and oil price increases rises sharply — and with it the risk that the inflation improvement of May PCE proves temporary rather than durable. The AI story and the Iran story are not separate. They are linked through oil prices, inflation, interest rates, and the discount rate that applies to every future earnings projection in the technology sector. Understanding that link is essential to understanding why a week that produced the greatest earnings report in semiconductor history also produced the Nasdaq's worst weekly performance in four months.</p>