Semiconductor shares bounced sharply in premarket trading after weeks of heavy selling, with Micron Technology surging 4.7%, Advanced Micro Devices rising 3.11%, Intel gaining 2.48% and Nvidia adding 1.16%. The move signals that bargain hunters are beginning to absorb the AI hardware and memory names that were dumped in last week’s rout, reigniting a familiar debate: has the latest chip sell-off opened another contrarian buying window?
Throughout this artificial intelligence cycle, steep semiconductor drawdowns have repeatedly proven short-lived. Earlier this year, cautious guidance from Broadcom erased more than a trillion dollars in sector value within days, only for the stocks to snap back once investors remembered that AI infrastructure spending was still accelerating. Global chip sales hit a record in 2025 and are projected to grow further in 2026, with AI-related silicon accounting for roughly half of the total. As long as data-centre build-outs remain insatiable and supply stays tight, every deep pullback looks cheap in hindsight.
For contrarian investors, the clearest signal comes from Micron’s management. On its latest earnings call, the company stated that the memory chip market is expected to remain “tight” beyond 2027, and it is locking in long-term supply agreements with downstream buyers. A prolonged supply crunch, reinforced by multi-year volume and price commitments, eases the fear of an imminent cycle reversal and makes the recent drop of more than 25% look increasingly attractive for those willing to step in.
Broadcom (AVGO) – A blue-chip in custom AI silicon, Broadcom designs specialised chips for Google, Meta and Anthropic, while also holding strong positions in data-centre networking and infrastructure software. Revenue is expected to jump 66% year on year in the current financial year, and the forward price-to-earnings ratio based on the fiscal year starting in November has fallen below 20. The stock is trading about 25% below its highs, and the average analyst price target of $510 implies roughly 38% upside from current levels. The key risk is customer concentration: if a major client such as Google or Meta scales back its custom-chip orders, revenue growth could slow materially.
Applied Materials (AMAT) – A leading manufacturer of chipmaking equipment, Applied Materials is a classic “picks-and-shovels” play on the AI boom. As foundries like TSMC and Intel keep expanding capacity, the company should enjoy strong order momentum. Analysts are forecasting 28% top-line growth for the next financial year. At around $530, the stock is down approximately 28% from its 2026 highs and trades at about 32 times next year’s expected earnings — not cheap, but still reasonable against a high-growth backdrop. The main concern is that revenue is highly dependent on the capital expenditure budgets of a small number of chip manufacturers; any sudden pullback in industry spending would weigh heavily on performance.
Micron Technology (MU) – A more cyclical bet on memory, Micron is benefiting from surging demand as generative AI models require vast amounts of storage, processing and data access. Revenue is on track to more than triple in the current fiscal year ending in August, and Wall Street expects a further 81% increase next year. While the cyclical nature of memory cannot be ignored, management’s guidance that tightness will persist until at least after 2027 provides a longer runway. The valuation looks starkly attractive: the stock trades at just 11.6 times expected fiscal 2026 earnings and 5.7 times fiscal 2027 estimates, with some analysts targeting a price as high as $1,500. The biggest danger is that if demand suddenly falters or supply catches up, memory prices could crash, so this position demands especially careful risk management.
History shows that buying the dip tends to work well while the upcycle remains intact, but cycles always turn eventually. Today’s supply-demand dynamics still support the medium-term thesis for chip stocks, and accumulating quality names on weakness is a rational strategy. Yet investors should approach this with a probability mindset, size positions prudently, and leave enough room to be wrong — because when the cycle does shift, the correction tends to run far deeper than 20%.
