Micron Technology (MU)

Total return at 5 years: -10.4%/year = -10.5% appreciation + 0.1% dividend. The target price ($569) is ex-dividend; the $3 in dividends collected over 5 years are added separately.

By both methods, the value today (DCF $315 · Multiples $460) is below the market price ($989).

Today — expensive, no margin of safety: at $989 trades ~115.2% above its value discounted to today (~$460); the expected return does not even reach the risk-free rate (4.5%).

At 5 years — Sobrevalorado: the expected total return is negative — the price already discounts a demanding scenario that, if not met, results in a loss.

The bridge: the return at 5 years falls below the risk-free rate (4.5%) — which is why there is not even a discount to today's value. To require a 15% annual return, it would need to be bought at ~$285.

Educational / informational. Does not constitute investment advice.

The inverse approach: instead of projecting growth to obtain the value, the market price ($989) is taken as given and it solves for what annual owner-earnings growth would need to hold for 5 years for the present value —at the method's rate (4.5%, no-growth terminal)— to equal that price. It is the disconfirmation test: the expectations the price already pays for, contrasted against the method's projection.

The price discounts growth (34.5%/year) above our base case (3.0%/year) → it is priced for a demanding scenario and leaves little cushion against a slowdown.

That growth implies ~$52.9 bn of owner earnings in year 5 (vs ~$13.9 bn of our base case). It recalculates if the DCF assumptions are edited.

Year-by-year projection of the selected scenario. From each year, two versions of the flow are derived: growth FCF (operating flow − total capex, the cash surplus) and maintenance FCF (the owner earnings: what the business yields if it only sustains its capacity). Cash accumulates the retained surplus —what is not returned as dividend or buyback—, so that EV falls and multiples compress going forward. The valuation is done on Utilidad neta (P/E mid-cycle). In edit mode, revenue, margins, capex, and exit multiples can be adjusted.

Valuation in two parts. The first is the business at market prices: the path describes Micron as if it sold its entire volume at the current price. Year 0 is fiscal 2026, closed on 3-Sep-2026 and 53 weeks long (revenue US$133,188 million, net income US$84,969 million, gross margin 80.7%), a cycle peak. Year 1 starts from the first-quarter fiscal 2027 guidance (revenue of US$61,500 million ± US$1,500 million, gross margin of about 85.95%, operating expenses of about US$2,310 million and earnings per share of US$37.84 on about 1,150 million diluted shares; release of 30-Sep-2026). From there quarterly revenue falls geometrically, 5.5% per quarter, to the mid-cycle level of the fifth year (US$92,100 million), and the net margin converges to the mid-cycle margin (27.3%); the extra week of year 0 is not extrapolated. The terminal value capitalizes that mid-cycle net income (US$25,100 million) at 13 times, never the peak. The earnings are operating profit after taxes, at the effective rate of fiscal 2026 (14.8%: US$14,761 million on US$99,671 million before taxes) and without interest on cash, which is added separately. The gross margin of each year is the operating margin plus operating expenses at the guided first-quarter pace (US$2,310 million per quarter). Year 1 capital expenditure is the minimum the company guides (about US$25,000 million in the first half, net of public incentives, and «higher» in the second: at least US$50,000 million), and later years fall with revenue to 29% at mid-cycle; working capital is assumed stable. Cash accumulates: it went from US$11,940 million to US$73,450 million in the year, so the flow is not entirely redeployed in fabs. 97.7% of free cash flow is retained (a repurchase of US$650 million and a dividend of US$0.60 per share, against free cash flow of US$58,963 million). The second part is the contracts with a price floor. During fiscal 2026 Micron signed multi-year mandatory-purchase contracts, mostly with the price inside a band with a minimum and a maximum. At 3-Sep-2026 the revenue committed at the minimum price totaled about US$134,000 million: about one quarter is recognized in the next twelve months and all of it within five years (10-K, revenue note). The split across years 2 to 5 is not published and is divided into equal parts; the five-year value does not depend on that split. The company expects that, even at the minimum price, these contracts will give a gross margin «meaningfully above» the peak quarterly margin of any prior cycle, which was 61.0% in the fourth quarter of fiscal 2018 (XBRL); that 61.0% is taken as the floor. The contract adds value only in the year when the path's gross margin falls below the floor: that excess, after taxes, goes to cash and is valued at 1× as a separate piece. It does not enter the multiple, because the contracts expire within the horizon and their renewal is not signed. What they yield above the minimum price is not credited, because the filing does not quantify it. Customer deposits (US$12,895 million, returned between 2029 and 2031 if the customer meets its minimum purchase) are subtracted as debt together with financial debt (US$5,179 million). The adverse case assumes the cycle turns after the guided quarter: revenue falls to US$68,000 million in the third year with a gross margin of 28% and stabilizes at US$65,000 million, at 10 times; it is the scenario where the floor of the contracts contributes most. The favorable case sustains the pace of the guided quarter for one year and converges to US$140,000 million with a net margin of 38%, at 15 times. Watch: the turn of the cycle and the renewal of the contracts.

Today's elevated multiple is the price of growth: if the business grows, the entry point cheapens on its own going forward (EV falls as cash increases). The exit multiple at 3 years is higher than the terminal at 5 years —at 3 years there is more growth still ahead—, so the value curve shows whether value creation is concentrated in the early or the later years. The required return is applied to the base scenario.

They are valued separately, with their own rationale, and are not incorporated into the base or the verdict (they are excess return). When assigning them value — in Editmode —, the total with optionalities updates live, without moving the base.

The verdict, the base CAGR, and the margin of safety are always calculated on the base; optionalities do not alter them (with optionalities at $0 they do not move).

Each card fixes a required annual return and answers: if the business is worth $569 in 5 years, what is the maximum that can be paid today to obtain that return? Since it now trades at $989, the margin of safety is how much cheaper the market is than that maximum. The three thresholds: 4% covers inflation (the floor), 10% is the long-term average return, and 15% is the level of a great investment.

With a target price of $569 in 5 years plus $3 of dividends collected (the dividend adds to the return, not to the price) and a required return of 4.5% annually, the maximum to pay today is $460. Against the current market price ($989), the margin of safety is -115.2% (trades above the maximum → a premium is paid) and the total return at that price would be -10.4% annually.

• Mid-cycle P/E, not peak. Valued by P/E on mid-cycle earnings, never the fiscal 2026 peak (US$84,969 million). Valuing the peak would be compounded optimism.
• The single-digit P/E deceives. The single-digit P/E is on peak earnings → the cyclical value trap. On mid-cycle earnings the stock trades at ~45×: expensive, not cheap.
• The DCF converges. The DCF (at the risk-free rate, on mid-cycle generating power) gives a present value far below the price: the normalized flow does not support the quotation.
• Margin of safety at the end. At ~$989 the base return (~-10%) does not meet the required bar: only the favorable scenario (return of ~-10%) slightly exceeds the price. There is no margin of safety.
• The balance sheet cushions. The cash of US$73,453 million, net of debt and customer deposits, protects the downside at the turn of the cycle, but does not justify the price on normalized earnings.
• Sensitivity — the cycle. The value is hypersensitive to where mid-cycle is placed: if artificial intelligence is structural, the level rises; if it is another peak, it reverts toward 2023. The contracts with a price floor reduce that dispersion only downward and only for five years.

Each margin over sales, year by year: historical (solid line) → projection (dotted).

• ✓ ROIC exceeds the cost of capital (~10%)
• ✓ High incremental ROIC (compounding)
• ! CFROIC backs up the ROIC (62%, cash vs. accruals)
• ✓ Healthy balance sheet (low corporate debt)
• ✓ Durable competitive moat (multiple advantages)

• Strong and predictable cash. Neither strong nor predictable across the cycle: free cash flow was −US$6,117 million (fiscal 2023), US$121 million (2024) and US$1,668 million (2025), thin even in a good year because of capital expenditure. Fiscal 2026 (US$58,963 million) is the exception, not the norm.
• Return on capital. Across the cycle it barely touches the 10% bar (the troughs and capital intensity destroy it); the mid-cycle level lifted by HBM and the contracts holds it above, without the durability of a wide-moat business.
• Creates economic value. At the peak, enormously; through the cycle, marginally — the business has destroyed value at the troughs (losses in the billions). Value creation depends on where in the cycle it is measured.
• The peak-P/E trap. The single-digit P/E is on peak earnings: the classic cyclical value trap. It is valued on mid-cycle earnings, never the fiscal 2026 peak.
• Capital intensity. Capital expenditure was 23% of revenue in fiscal 2026 and 42% in 2025; the business consumes cash to maintain bit share, a structural drag on returns.
• Distortions to watch. (1) Inventory charges at the troughs; (2) CHIPS Act subsidies that offset capital expenditure; (3) peak earnings that make a P/E look cheap; (4) customer deposits, which are debt to be returned between 2029 and 2031 and not revenue.

Annual levels from the official filing (10-K); the % over each bar is the year-over-year (YoY) growth vs the prior year.

Revenue is memory whose price is set by the industry: volume of bits times price per bit, in DRAM (76%) and NAND (24%). The margin depends on selling prices and on the reduction in cost per bit. In fiscal 2026 DRAM grew 252% with an average price 180% higher, and the cloud unit (HBM, high-capacity modules and low-power server DRAM) reached US$43,085 million. Capital expenditure, at US$30,712 million, is the counterpart: extremely capital intensive, with CHIPS Act support. These are peak-of-cycle levels, not a sustainable rate.

The % are the annual (year-over-year) growth: each year —historical and projected— vs the prior one; the TTM (trailing 12m) vs the TTM of a year ago, to avoid overlapping windows. The historicals are exact figures from the official filings; the projected years come from the year-by-year model (the intermediate years without their own series are interpolated between the anchors). The projected columns (+1y…+5y) are 12-month windows counted from the TTM close (3-sep-2026): the projection starts from the most recently reported data, not the fiscal year. The projected base is realistic and unbiased — the risk discount is applied at the end, via the required return. The rationale for each metric is in the (i).

• Real growth, not nominal. The +256% of fiscal 2026 is cyclical and driven by price (DRAM average price +180%, NAND +200%), not structural volume: it cannot be extrapolated.
• Where the engine comes from. DRAM (76%) and the cloud unit (US$43,085 million, +219%) through artificial intelligence. The engine is data center memory, the differentiated segment.
• Driven by volume and price. In fiscal 2026, price (+180% in DRAM) far more than volume (bits in the mid-20% range). Peak-of-cycle growth is intrinsically unsustainable.
• Realistic deceleration. The base starts from the first-quarter fiscal 2027 guidance and normalizes toward mid-cycle earnings as prices revert, with quarterly revenue falling geometrically.
• Reinvestment runway. Wide in HBM/node, but at historically meager returns and with punishing capex; bit growth demands constant investment just to maintain share.
• Durability of AI demand. The company states that demand exceeds industry capacity and signed mandatory-purchase contracts over several years. If it is structural, the mid-cycle level rises; it is the key variable, still unproven across a downturn.

• Relevant in 10 years / recurrence. Yes as a category: memory is a structural input to all computing and artificial intelligence expands it. Contractual recurrence is new and limited: the strategic contracts of fiscal 2026 commit about US$134,000 million, but they expire within five years and the rest of sales are at market prices.
• Market share / cost. Third in DRAM, in an oligopoly. Node leadership gives a competitive cost, but it is a price taker, not a price setter.
• Pricing power. Null in the structural sense: the price is set by industry supply and demand. The average selling price of DRAM rose about 180% in fiscal 2026 because of the cycle, not because of Micron's pricing power; the contracts set bands with a minimum and a maximum.
• Operating leverage. Extreme and double-edged: the gross margin jumped from −9% (fiscal 2023) to 80.7% (fiscal 2026); at the turn it collapses with the same violence, on a price Micron does not control.
• Capital cycle / competition. Capital enters where returns are high and compresses them: the large producers are shifting capacity to HBM and the Chinese entrants (CXMT, YMTC) add supply, the signal of the next turn of the cycle.
• Behavior in a recession. Highly cyclical: memory demand falls with technology capital spending. The 2023 trough (a −$5.8bn loss) shows the amplitude; AI HBM could cushion the next one, with no guarantee.

Not all debt is equal: only the structural needs refinancing; the rest is operational (self-liquidating).

Structural debt is what is exposed to the contraction phase of the cycle; operational debt (leases, matched funding) self-liquidates with the business.

• ✓Leverage (net debt / EBITDA)Net cash $55.4 bn
• ✓Interest coverage (EBIT / interest)235.8x
• –Liquidity (current ratio)no data
• !Cash quality (CFROIC vs ROIC)CFROIC backs 62% of ROIC
• !Value creation (ROIC − 10% bar)+2pp
• !Malinvestment test (capex vs incremental ROIC)Strong investment mode 2.0x
• !Float / working capitalConsumes cash $3 bn (positive WC)
• ✓Dilution (SBC % of revenue + shares)SBC 1.0% of revenue

A traffic-light interpreted by the method (not generic): float (negative WC) adds up, capex is judged by incremental ROIC (malinvestment test), and a lender is not subjected to corporate solvency. The (i) shows the derivation of each number.

• Net cash. Cash and investments of US$73,453 million against US$5,179 million of financial debt and US$12,895 million of customer deposits → net cash of US$55,379 million; it repaid US$10,040 million of debt in fiscal 2026. Cash-rich balance sheet.
• Ample coverage. Interest expense is trivial (US$106 million) against operating cash flow of US$89,675 million; at cycle troughs coverage narrows, but debt is low.
• Capital intensity. Capital expenditure (US$30,712 million in fiscal 2026, with guidance of about US$25,000 million in the first half of 2027 alone) is the greatest cash risk: at the cycle trough it exceeds operating cash flow.
• Cash generation. Thin across the cycle because of capital expenditure; the current peak generates enormous cash (free cash flow of US$58,963 million), but it is the exception.
• The cycle, not solvency. The risk is the cycle (memory prices), not the balance sheet: with net cash and CHIPS Act support, Micron rides through the troughs without solvency risk.

Who owns the shares — the alignment and whether there is a controlling shareholder.

• Capital allocation. Fab capital expenditure dominates; capital return is subordinate (token dividend, repurchase of US$650 million). Right for the phase, but at historically meager returns.
• Alignment / skin in the game. Professional management, not a founder: insiders below 1% of capital. Holding guidelines (6 times salary), but without the alignment of a large economic ownership.
• Return of capital. Dividend of US$0.15 per quarter (US$0.60 annualized, a yield of ~0.06%) and a repurchase of US$650 million in fiscal 2026; authorization expanded to US$35,160 million from December 9, 2026. The shareholder depends on appreciation, not on cash return.
• Integrity / communication. Communicates the cycle transparently (quantifies the contracts, the deposits and the letters of credit); the chief executive has deep operating knowledge of memory.

• The mandatory-purchase contracts with a minimum price and the first-quarter fiscal 2027 guidance (revenue of US$61,500 million) give visibility to a peak that the market may be capitalizing as if it were the normal state; the method does not capitalize it.
• The P/E on peak earnings is single-digit and makes the stock look cheap, but it is the signature of the cyclical value trap: a low multiple on earnings that the cycle can reverse. On mid-cycle earnings the stock trades at ~45×.
• The cash of US$73,453 million and the leadership in HBM feed the quality narrative, but they do not change the fact that memory is a product whose price is set by the industry, with a cycle trough that three years ago was a loss of US$5,833 million.

The perception (a peak sustained by contracts, a single-digit P/E) is better than the mid-cycle reality. Without a favorable gap between perception and reality there is no margin: the price discounts the peak as if it were the normal state.

The annual return (CAGR at 5 years) in each scenario, with the total period return below — the margin of safety made visual: upside range wide, downside range narrow.

Even in the bear scenario, the return holds at -23%/year (-72% total): the margin of safety protects the downside. The bull (+2%/year, +10% total) exceeds it comfortably — a favorable asymmetry, with a narrow downside range and a wide upside range.

• The cycle turning (oversupply). The central disconfirmer: supply rebalances (the large producers and China) and prices fall, as in 2023. Peak earnings collapse; the floor of the contracts cushions only for five years.
• HBM competition. Samsung and SK hynix compete for artificial intelligence memory; if Micron loses share or the premium margin compresses, the differentiator erodes.
• Chinese entrants. CXMT (DRAM) and YMTC (NAND) add supply and litigate against Micron; they are the signal of the capital cycle that compresses returns.
• Contract counterparty. Minimum purchases depend on customers fulfilling them. The deposits of US$12,895 million and the letters of credit for US$7,000 million mitigate the risk; if the market price exceeds the maximum of the band, the contracts cap the upside.
• CHIPS Act and China. The subsidies (up to US$6,400 million) are subject to conditions and clawback; the ban by China's cybersecurity authority is a persistent drag on demand; there are cease orders in Germany over NAND patents.

Each thinker's analytical framework applied to our data.

• Buffett / Graham Quality + margin of safety

  A narrow moat and ROIC 12% above the 10% bar, but the price sits above value (no margin) → great business, expensive.

• Peter Lynch Growth at a reasonable price (GARP)

  A stalwart growing 3% at a PEG of 4.5 → expensive for its growth.

• Joel Greenblatt Cheap and high-return (Magic Formula)

  Earnings yield 2% (EBIT/EV) + ROIC 12% → falls outside the Magic Formula.

• Howard Marks Perception vs reality + cycle

  The price discounts 35% vs our 3%: priced for perfection, no favorable mispricing.

• Seth Klarman Capital protection (bear scenario)

  Bear-scenario floor -23%/yr over 5y (material loss) → risk of permanent capital loss.

• Pat Dorsey Moat strength (Five Rules)

  A narrow moat, stable; sources: cost advantage, efficient scale, intangibles, switching costs → partially passes the Five Rules.

• Aswath Damodaran Expectations implied by the price

  The price demands 35% growth, far above our 3% — heroic assumptions.