Revenue is up 379%, gross margin is 87% and the FQ1 guide is $61.5B, yet the shares trade 30% below the June high at ~5x forward earnings. DRAM forgets everything when the power is cut; investors are pricing the chance that AI-capex power gets cut and these earnings go with it. Six analyst lenses, three scenarios, four horizons.
Gray = actual price into today ($84 end-2024 → $1,255 intraday peak Jun 25 ’26 → $193 intraday low Nov 21 ’25 → $877.55). Peak and low are intraday extremes; other points are daily/weekly closes, and the 2024–mid-2025 points are derived from dated % performance disclosures, so treat early history as approximate. Colored paths = synthesized scenario midpoints, probability-weighted (bear 30% · base 45% · bull 25%) on a log price axis. Street 12-month consensus ≈ $1,561 (49 analysts, “Strong Buy”).
Those probabilities are a judgment call — so make them yours. Drag to set how likely the bear and bull cases are (base takes the remainder); the blended target, the dotted line on the chart, and the prob-weighted row of the scenario cards update live.
One set of fundamentals, six frameworks. Each lens is a synthesized perspective with its own 12-month target (panel mean ≈ $1,088 vs a $1,561 Street consensus).
Revenue +379% and an 87% gross margin, with FQ1 guided to $61.5B and supply tight through FY28. Over 75% of 2027 output is already committed, so earnings are visible, not hoped for. At 5x forward EPS the stock does not need a premium multiple, only a normal one.
$62.3B of FY26 adjusted free cash flow is a ~6.3% yield on a ~$993B market cap, with $68.3B of net cash and buybacks unlocking Dec 9. EV/EBITDA is ~8.5x and trailing P/E ~11.8x, against 17.6x for SanDisk and 9.5x for Samsung on one screen.
Memory has produced 73%, 55%, 43%, 51% and 36% drawdowns this century, each from a “this time is different” peak. A 5x forward multiple is the market pricing peak earnings, not a bargain. Mid-cycle EPS near $120 at ~4x gives the downside.
Three suppliers dominate DRAM and HBM, and Micron now has 26 strategic customer agreements covering over 35% of revenue through 2030, with $12.75B of customer deposits received in FY26. Contracts turn a commodity into something closer to a capacity-reservation business.
Shares are 30% below the $1,255 high, down 19% over the recent leg, and trade at ~11.8x trailing earnings against ~27x at the June close. A partial reversion to ~6x FY27E is the trade; the trend is still broken until $1,030 is retaken.
On Oct 22 the SOX fell over 6% on AI-funding fears, rising oil and rates. Micron’s earnings are a derivative of hyperscaler capex, which depends on cheap capital. Until funding stabilizes the multiple stays near 4–5x.
Eleven selected firms, sorted low to high; the dashed line is today’s $877.55. Only one of the sampled targets sits below it.
Sell-side targets as listed Oct 1–23, 2026 (Google Finance, MarketBeat); consensus from StockAnalysis (49 analysts, Strong Buy). Aggregators disagree on the full range (one lists a low near $192 and high near $2,205, likely stale), so only the sampled names are shown. All sampled firms are Buy except JLPascual (Sell); no Holds appeared in the sample.
Synthesized scenario midpoints (mid-year), each EPS × exit multiple. Returns vs. today’s $877.55. These are illustrative frameworks, not forecasts; five-year outcomes hinge on whether the memory cycle is tamed.
Micron went from roughly break-even FCF to $62B in two years. The bull and bear cases live in the gap between the capex and FCF bars.
Actuals from Micron 8-K/10-K: revenue $25.1B/$37.4B/$133.2B; net capex $8.1B/$13.8B/$27.4B; adjusted FCF $0.4B/$3.7B/$62.3B; cash & investments $9.2B/$11.9B/$73.5B. FY27E: revenue $275.5B and FCF $142.9B are analyst consensus; capex ≈ $50B is my annualization of the ~$25B planned for H1 (not guidance); cash is not estimated. Micron is net-cash (debt only ~$5.2B after $10B of FY26 repayments), so cash replaces the debt bar. Fiscal years end late Aug/early Sep.
Each scenario is an EPS estimate times an exit multiple. Here is the earnings ladder the base case is built on.
Reported: FY25 $8.29, FY26 $75.52 (GAAP $74.33). FY27E $176.55 and FY28E $206.89 are Yahoo Finance consensus (35–36 analysts, Oct 2026; other aggregators range from ~$161 to ~$184 for FY27). FY29E–FY31E are the base-case path (cycle normalization, then buyback-driven recovery), not consensus; older consensus surveys show FY29 EPS falling toward ~$125, which is the market’s cycle fear in one number. Base 5-yr $1,400 ≈ ~$200 × 7x.
Latest fiscal-year and Q4 FY26 year-over-year growth. Read these against a stock 30% below its high.
Revenue and DRAM are olive; the faster profit and data-center lines are clay. Sequential growth was also +31% in Q4 and gross margin reached 87%. That gap, record results against a falling price, is the bull case in one chart; the bear answer is that growth rates this extreme are what cycle peaks look like.
The entire valuation argument compresses into one disagreement: are these earnings durable memory or volatile memory?
Over a 3–5 year horizon. The hot corner, a likely and high-impact price roll-over, is the one that matters; the next tier is mostly “possible.”
Memory prices stop rising and then fall; gross margin of 87% resets and EPS follows. Every prior cycle ended in a 36–73% drawdown.
Samsung, SK Hynix and CXMT (4F2 DRAM, large capital raises) add bits faster than AI demand absorbs them.
Meta’s first negative free cash flow since 2012 and Huang’s “capital is scarce” remark raise the odds that customers slow orders.
About $25B of capex is planned for H1 FY27 alone, nearly all of FY26’s $27.4B; it caps how much free cash flow reaches buybacks.
Beta above 2 and daily moves of 4–6% make drawdowns routine; the stock fell 25% in the four months after the June high.
A few AI platform buyers drive the order book; one pause hits revenue and pricing together.
Strategic customer agreements may bend if spot prices fall well below contract levels.
Higher incentive compensation and start-up costs already weigh on FQ1 gross margin guidance (about 86.25%).
A Taiwan supply disruption or export-control shock would hit HBM and DRAM output and reprice the sector overnight.
Hover the dotted terms in the metrics, or scan the desk’s working definitions here.