All-In on Micron: Can a Low Valuation Deliver a Rerating of the Memory Business?

Analysis

The OP views Micron (MU) as an asymmetric-return “memory bet,” using Pascal's wager as an analogy for going all-in: if Micron remains a traditional cyclical stock, its current valuation already reflects a future downturn, so losses should be limited; if AI demand, high-bandwidth memory (HBM), and multiyear take-or-pay contracts change the nature of the business, the company could command a valuation approaching that of an AI infrastructure company, generating severalfold returns. The argument begins with approximately 14 times trailing twelve-month earnings and approximately 6 times forward earnings, then interprets those low multiples as evidence that the market has already priced in falling prices, margin compression, and an earnings collapse. The author then proposes that memory could shift from a cyclical commodity to infrastructure, using NVDA, MRVL, and AVGO as valuation benchmarks and another comparator referred to as “SKHY” to suggest similar rerating potential. These names are merely analogies, not independent assessments of those stocks in the post, so MU is the only core affected ticker. The post claims that the share price could easily reach 3 to 4 times its current level, but provides no explicit dollar price target, time horizon, projected earnings per share, comparable-company multiples, or complete valuation model. Assuming earnings forecasts remain unchanged, deriving a share price of 3 to 4 times the current level from a forward P/E of 6 implies that the multiple would need to rise to approximately 18 to 24. This is only an arithmetic reconstruction of the argument, not a target valuation supplied by the author. Regarding the position, the author reports a 100% allocation to MU, holding approximately 126 shares, and adds in the comments that the entry price was approximately €945, implying a nominal cost of approximately €119,100. No trade confirmations, exchange rate, account size, information on margin use, or holding period are provided, so “100%” cannot be equated directly with the author's entire personal wealth. The post has no image attachments, financial-statement charts, contract screenshots, or position screenshots to substantiate its claims. Potential catalysts include continued realization of AI-related demand, expansion of the HBM business, improved earnings visibility from multiyear contracts, and market acceptance that Micron's business has undergone a structural change, but the main post identifies no specific event dates. The author's implicit assumptions include reliable and sustainable forward earnings, contracts covering enough of the business and constraining prices and purchase volumes, margins avoiding another sharp decline, and a capital-intensive memory manufacturer receiving valuations comparable to the listed AI companies. The most important flaw is treating a low P/E as downside protection: even if the market already fears a cyclical downturn, further downward revisions to earnings forecasts could still cause a substantial share-price decline, and the low multiple may simply be an artifact of peak-cycle earnings in the denominator. The post also compresses the outcomes into two scenarios, omitting intermediate paths such as slowing demand, supply expansion, product competition, and customer capital-expenditure cuts, without calculating probabilities, stress scenarios, or the magnitude of permanent losses. Overall, it is closer to a concentrated-bet narrative dependent on a valuation rerating than a risk-reward analysis that has demonstrated “limited downside.”

Discussion Summary

Disagreement remains over whether AI can change the memory cycle. Supporters believe tight supply and demand could persist through 2027—2028, that lagging earnings estimates make valuations look low, and that the argument that “a rapid rise must lead to a fall” reflects price anchoring. Opponents emphasize that a low P/E at the cycle peak does not protect against downside and that the risk of going all-in far exceeds that of a broad-market drawdown. New discussion focuses on order cancellations, capital expenditure cuts, and declining selling prices. Someone corrected an overstated calculation of the impact of selling prices halving on gross profit, but this did not resolve the debate over earnings pressure. On Chinese competition, one side believes security restrictions protect the US market, while the other argues that substitution in overseas markets would still depress global prices; both are community assessments. OP previously added an entry price of approximately 945 euros but offered no substantive response this time and still has not explained the risk limits of the all-in position.

Fact Check

• supported: MU currently trades at approximately 14 times trailing twelve-month earnings and 6 times forward earnings. — As of the close on October 9, 2026, the market-data source lists a trailing twelve-month P/E of 13.84 and a forward P/E of 5.84, consistent with the post's approximate figures. The forward multiple still depends on earnings forecasts and cannot be treated as a guaranteed return. ([stockanalysis.com](https://stockanalysis.com/stocks/mu/statistics/?utm_source=openai))
• supported: Micron has multiyear take-or-pay contracts that can improve business visibility. — The fiscal 2026 10-K confirms the existence of multiyear take-or-pay agreements with contractually binding purchase volumes. Most strategic agreements include price floors and ceilings, and customers provide cash deposits and letters of credit. However, some contracts remain short-term, and the existence of contracts does not mean all revenue and profits are protected. ([sec.gov](https://www.sec.gov/Archives/edgar/data/723125/000072312526000023/mu-20260903.htm))
• questionable: AI, HBM, and long-term contracts have made Micron no longer a purely cyclical company. — The company discloses that it began volume production of HBM4 in 2026 and believes strategic agreements can improve the stability and predictability of financial performance, supporting the direction of an improved business structure. However, these disclosures are insufficient to prove that cyclical risk has disappeared, and the company continues to disclose intense competition and the risk of competitors cutting prices. ([sec.gov](https://www.sec.gov/Archives/edgar/data/723125/000072312526000023/mu-20260903.htm))
• unsupported: A low P/E means a cyclical downturn is fully priced in and losses will be limited if the thesis is wrong. — This is an inference about market pricing and the limits of risk, not a fact that can be established directly from a P/E ratio. The main post provides no normalized earnings, margin stress tests, or valuation-floor model. If earnings forecasts decline, the current low multiple cannot guarantee limited share-price losses.
• unsupported: If the structural transformation holds, MU can easily reach 3 to 4 times its current share price. — The main post provides no target time horizon, earnings trajectory, or comparable-valuation calculations. With earnings unchanged, a share price of 3 to 4 times the current level from a P/E of 6 requires a valuation of approximately 18 to 24 times earnings, but neither the reasonableness nor the attainability of that multiple has been demonstrated.