DA Davidson raises Micron stock price target on memory market shift

The key takeaway is not just that memory is re-rating, but that the market is beginning to assign structural scarcity economics to a category long treated as cyclical and commoditized. HBM tied into AI server architecture creates switching costs and qualification friction, which should extend gross margin durability for the memory leaders and compress the valuation gap versus CPU vendors over the next 12-24 months. That supports MU, but it also raises the odds that the entire AI hardware stack stays capital intensive longer than consensus expects, benefiting equipment and packaging suppliers while eventually pressuring hyperscaler ROI.

The second-order effect is that if memory starts trading more like a strategic component, pricing power migrates from end demand to supply discipline. That is bullish for vendors with the cleanest product mix and the strongest customer lock-in, but it also makes the cycle more fragile: a modest digestion in AI capex could produce a much sharper de-rating than the current multiple implies because the stock is already discounting years of exceptional growth. The risk window is 3-6 months for sentiment reversal, but 12-18 months for fundamentals to prove whether this is a durable regime change or simply a shortage-driven squeeze.

The contrarian view is that the market may be extrapolating HBM scarcity far beyond the point where capacity comes online and customer bargaining power returns. If AI deployments shift from training to inference-heavy workloads, memory intensity could normalize faster than bulls expect, especially if hyperscalers optimize around cost per token rather than raw bandwidth. In that scenario, MU still wins on earnings, but the multiple expansion becomes the vulnerable part of the trade.