Prediction: 1 Semiconductor Stock That Will Be Worth More Than Micron by 2028

Micron has surged nearly 800% over the past 12 months and briefly joined the $1 trillion club, driven by severe HBM/DRAM shortages and pricing power, including a mid-60% sequential DRAM price increase last quarter and 207% year-over-year segment revenue growth. The article argues the memory cycle should peak in mid-2028, with analysts estimating FY2027 EPS of $106 before a modest decline in FY2028 and a potentially much lower valuation by 2028. It also highlights ASML as a better long-term beneficiary of semiconductor capacity expansion, with management guiding 2026 revenue of 36B-40B euros and investor-day 2030 revenue targets of 44B-60B euros likely to be exceeded.

Analysis

The key second-order issue is not whether memory is strong today, but whether capital intensity will outrun pricing power before the cycle normalizes. HBM demand is structurally real, yet it also incentivizes a synchronized supply response from DRAM peers and equipment vendors, which usually compresses returns faster than sell-side models expect. In that setup, MU can still post spectacular near-term earnings, but the equity can de-rate well before absolute profits roll over if the market starts discounting 2028/2029 supply relief.

ASML is the cleaner way to express the same AI capex theme because it monetizes capacity expansion across the industry rather than a single memory sub-cycle. The underappreciated benefit is that even if memory pricing cools, foundry and memory fabs still need to refresh toolsets and expand install bases, supporting a longer revenue runway than chip ASPs themselves. The counterpoint is valuation: the market already knows ASML is the toll booth, so upside depends on sustained capex acceleration, not just a good order book.

The contrarian read is that the market may be overpaying for peak-cycle earnings in MU while underappreciating how quickly consensus can move when capacity catches up. This is a classic moment where reported growth is strongest near the top, which tempts investors to extrapolate just as forward margins begin to flatten. The bigger setup is relative value: long the infrastructure enabler, short the cyclical beneficiary that is most exposed to mean reversion.

NVDA remains a modest beneficiary through HBM demand, but the incremental upside there is likely less asymmetric than in prior AI waves because memory scarcity can become a tax on system deployment rather than a pure demand booster. INTC is only a marginal passerby in this trade, but any broad semiconductor capex upcycle helps sentiment around domestic manufacturing and foundry ecosystem narratives. NFLX is effectively irrelevant here except as a reminder that not all mega-cap momentum belongs in the same basket; this is a very specific capex-cycle, not a broad AI-beta trade.

Key Decisions for Investors

• Short MU on strength into the next 1-2 quarterly prints; target a 6-12 month horizon. Best risk/reward is a de-rating trade if the market starts pricing 2028 oversupply before earnings actually roll over.
• Go long ASML vs short MU as a relative-value pair for the next 12-18 months. The pair isolates AI capex expansion while reducing exposure to memory ASP mean reversion.
• Buy ASML call spreads or outright calls only on pullbacks tied to order fears, not after breakout moves. Use a 6-9 month tenor to capture 2026-2028 capex revisions without paying too much theta.
• For more convex exposure, consider a bearish MU put spread targeting post-earnings guidance resets over the next 1-2 quarters. The trade works if consensus begins to front-run 2028 supply normalization.
• Keep NVDA as a tactical long only on evidence that HBM bottlenecks are resolving without destroying GPU deployment volumes; otherwise, trim and rotate into ASML where cycle duration is longer.