To The Moon

On Thursday 6 November, $Lam Research(LRCX)$ fell 16.4% as the Philadelphia Semiconductor Index dropped 12.25% to 9,011.14, its steepest one-day decline since March 2020. Before Friday’s open, Lam’s board authorized a $25 billion buyback with no termination date, on top of what remained of its prior program. The stock closed Friday at $199.48, up 7.71%, against 4.21% for the same SOX.

A buyback that size is a board putting capital behind a view of demand. The question is narrow and testable: did the board buy because it knows something the tape does not, or because a stock at half its June price was too cheap to ignore while 2027 capex stays unproven?

Figure 1 — The four numbers that frame this piece: buyback size, the Thursday move, the Friday close, and the cash that funds it. Sources: Lam Research release via MarketDash, 6 Nov 2026; Investing.com; Barchart.

Two days, two verdicts on the same set of facts

Thursday’s selling was not company-specific. Post-midterm signals on data-center regulation, layered on a payrolls print that pushed December rate-hike odds to 78% (LSEG, via BigGo Finance), hit every name with AI capex exposure. Equipment, memory and NAND were sold as one instrument.

Figure 2 — Thursday’s session: the whole AI hardware chain repriced together, equipment down nearly as much as memory. Sources: Investing.com citing LYNX Equity Strategies, 6 Nov 2026.

$SanDisk Corp.(SNDK)$ lost more than 22%, $Micron Technology(MU)$ 17.5%, $Advanced Micro Devices(AMD)$ 17.3%, and Lam 16.4%. A 16% move in a company with $8.22 billion of quarterly revenue is a statement about the cycle, not the quarter Lam had just reported. The market was pricing a 2027 wafer fab equipment number nobody sees until January.

Why Lam got the extra four points

Friday separated the group. Lam gained 7.71%, $Applied Materials(AMAT)$ 6%, $ASML(ASML)$ 5%, and the SOX 4.21%. Applied and ASML had no news of their own. Lam did.

Figure 3 — Friday’s rebound, ranked: Lam’s outperformance versus equipment peers is the market pricing the authorization, not the sector bounce. Sources: Barchart (LRCX close); TIKR via Finnhub (AMAT, ASML); TradingKey (SOX).

The size matters more than the announcement. Lam generated a record $3.24 billion of free cash flow in the September quarter, so $25 billion is roughly two years of cash — and the authorization obliges it to buy nothing. This is not a company signalling acceleration. It is a board underwriting a floor: we can return this much and still fund the plan, because the cash generation we see does not resemble a demand collapse.

CFO Doug Bettinger framed it as confidence in cash generation and in Lam’s position at key technology transitions. LYNX Equity’s KC Rajkumar read it sharper in a note titled “AI Semis: When There Is Blood on the Streets”: the timing suggests the demand deterioration implied by the share price is not showing up in the business, and memory and flash shortages still insulate suppliers from the financing worries now attached to hyperscalers and neoclouds.

The guidance arithmetic nobody quoted

Here is the part the “chip stocks bounce” coverage skipped. Lam’s September quarter beat, and its December guide brackets consensus rather than breaking it.

Revenue came in at $8.22 billion, up 52% year on year, against $8.15 billion expected, with an adjusted gross margin of 52.4%. Adjusted EPS was $2.20 versus $2.18. For the December quarter, Lam guided revenue of $8.2–9.0 billion against an $8.53 billion consensus, and adjusted EPS of $2.18–2.48 against $2.31.

Figure 5 — FQ1 actuals against the FQ2 guide and consensus: the midpoint sits above consensus on both lines, and the low end sits below it. Sources: Lam Research FQ1 results and FQ2 guide via MarketDash, 6 Nov 2026.

Read the guide’s shape, not its midpoint. A range spanning from below consensus to 5.5% above it describes a management team that cannot see the December quarter clearly, not one that sees a cliff. That fits a customer base still spending but unwilling to commit beyond a quarter — the defining feature of this cycle.

The bigger number is 2027. Management calls for wafer fab equipment spending above $200 billion next year, up from the mid-$150 billion range in 2026, and will update that forecast on its January call. Analysts, per TIKR, still expect Lam’s normalized EPS to more than double between fiscal 2026 and 2028. Those estimates did not move on Thursday; only the multiple did. Lam closed Thursday at 22 times forward earnings, under its five-year average of 24 times and far from June’s 57 times.

Bull Case

1. The buyback is funded from operating cash, not leverage: $3.24 billion of September-quarter free cash flow covers the authorization in about two years, and the program has no expiry.

2. Estimates are intact. No sell-side model cut fiscal 2027 revenue on the rout; the cuts were to price targets, which is a multiple decision, not an earnings decision.

3. Valuation has reset to the low end of its own history — 22 times forward against a 24 times five-year average — while the earnings base is guided up, not down.

4. Etch and deposition intensity rises at exactly the transitions that survive a capex slowdown: advanced logic and high-layer-count memory, where tool spend per wafer climbs even if wafer starts do not.

5. The signal is corroborated. KLA’s September-quarter revenue rose to $4.14 billion from $3.21 billion a year earlier and it guided December to $4.7 billion at the midpoint — a peer seeing the same order book.

Bear Case

1. Price targets were cut anyway, and by more than the stock’s Friday bounce. Needham went from $390 to $290, Morgan Stanley from $385 to $339, Mizuho from $365 to $360.

2. A buyback is not demand. $25 billion of repurchase authority says nothing about whether TSMC orders more tools in 2027; it says the board thinks its own shares are the better purchase.

3. The $200 billion WFE call is management’s own forecast, due for revision in January. If AI capex cools after the midterms, that number breaks first.

4. Memory is carrying this cycle. If the mix stays memory-heavy, Lam’s etch and deposition share of each dollar is lower than in a logic-led upturn.

5. The technical picture disagrees: Barchart’s own model carries a 72% sell reading, and the stock is still less than half its late-June level despite Friday’s gain.

Figure 4 — The counter-chart: three target cuts announced after the print, in percentage terms. The board and the sell-side looked at the same quarter and reached opposite conclusions. Sources: MarketDash analyst roundup, 6 Nov 2026.

Strategic Slam

$LRCX(Lam Research)$ is a cash-return story with a capex option attached, not a growth story. The position that makes sense gets paid while waiting: the buyback sets a mechanical bid under the shares, the dividend adds $0.33 a quarter, and the January WFE update either reprices the multiple toward 24 times or confirms the bears. What would make me wrong is not a bad quarter but a 2027 WFE number below $200 billion on that call. I would size accordingly, not before.

For the group, $KLAC(KLA Corporation)$ is the cleanest read on the same order book. $TSM(TSMC)$ raised 2026 capital spending to $60–64 billion in July and publishes its 2027 plan in mid-January — the best early read on whether Lam’s $200 billion exists.

Catalyst Calendar

The calendar is compressed: nothing to trade between now and January except the buyback itself.

Figure 7 — Dated catalysts: dividend mechanics in December, then three January events that decide the thesis. Dates marked expected are not yet confirmed by the companies. Sources: MarketDash (dividend dates); TIKR via Finnhub (TSMC timing, Lam January call), 6 Nov 2026.

What would change the verdict

Two numbers settle this. KLA’s December-quarter revenue against its $4.7 billion midpoint guide says whether the order book held through the political noise. Lam’s January WFE update says whether management still believes its own $200 billion.

Figure 6 — The four outcomes that matter. The bottom-right cell is the one that invalidates the buyback read entirely. Sources: TIKR via Finnhub (KLA guide, Lam WFE call), 6 Nov 2026.

Verdict

The board was right about the price and is unproven about the cycle. A 16.4% single-session decline in a company generating $3.24 billion of quarterly free cash flow was a sentiment event, and $25 billion of authorized repurchase is the right institutional answer. But the authorization stops short of what bulls want it to mean: it is not a demand forecast. Lam put a floor under its equity without committing to a 2027 number it cannot yet see. Net: constructive on the cash-return math, neutral on capex until KLA prints and TSMC publishes.

Key Takeaways

1. Lam authorized a $25 billion buyback with no end date within 24 hours of a 16.4% drop, and out-gained every equipment peer on Friday.

2. That authorization equals roughly two years of the $3.24 billion record free cash flow Lam generated in the September quarter.

3. FQ1 beat on both lines; the FQ2 guide brackets consensus, spanning $8.2–9.0 billion against $8.53 billion expected.

4. Consensus estimates did not fall — only price targets did, with Needham cutting from $390 to $290.

5. The thesis resolves in January: KLA’s December quarter, TSMC’s 2027 capex plan, and Lam’s WFE update.

The question I would put to the room: if the board will commit $25 billion at $199, what would 2027 WFE below $200 billion have to look like for that call to be wrong — and would you rather own the shares before or after you find out?

Tickers: $LRCX(Lam Research)$, $MU(Micron)$, $AMD(Advanced Micro Devices)$, $SNDK(SanDisk)$, $AMAT(Applied Materials)$, $ASML(ASML)$, $KLAC(KLA Corporation)$, $TSM(TSMC)$

Sources: Lam Research $25 billion repurchase authorization and $0.33 quarterly dividend, 6 Nov 2026 (company release, via MarketDash and Barchart); Lam Research FQ1 FY2027 results and FQ2 guidance (via MarketDash); Lam September-quarter free cash flow of $3.24 billion and 2027 WFE outlook above $200 billion (TIKR via Finnhub); Thursday 6 Nov session moves and SOX close of 9,011.14 (Investing.com citing LYNX Equity Strategies); Friday 6 Nov closes — LRCX $199.48 +7.71%, AMAT +6%, ASML +5% (Barchart; TIKR via Finnhub); TSMC 2026 capex of $60–64 billion and mid-January 2027 plan (TIKR via Finnhub); analyst target changes (MarketDash); forward P/E of 22x versus five-year average 24x (TIKR via Finnhub); December rate-hike odds of 78% per LSEG (BigGo Finance); Barchart technical reading (Barchart).

Disclaimer: This is analysis and opinion for information only, not investment advice or a recommendation to buy or sell any security. Figures are as reported by the sources listed and may be revised. I hold no position in any name discussed. Do your own work.