Huge bearish bets hit Micron and Nvidia as tech rebounds — buy the memory dip against the crowd

Verdict: memory fundamentals look real, but the entry trigger isn't — wait for confirmation

Over a 6-12 month horizon, the business case behind this idea is strong: Micron's latest quarter (ended September 3, 2026) showed revenue of $54.2B, up 30.8% quarter over quarter, with a 69.5% net margin and free cash flow up 87.1% to $32.9B, while Nvidia's CEO publicly called AI memory demand 'off the charts' (per CNBC, October 22). The strongest counterpoint is that provisional ownership filings for the period ending September 30, 2026 show net open-market insider selling at both companies — about $263M at Micron and $570M at Nvidia — and Micron's history of gross margin swinging from negative 33% to 87% makes today's record margins a peak-cycle risk, not a safety net. The attached trade rules can't help: the idea could not be backtested because raw and adjusted price series for both names use different bases, and its coded entry requires price below $0, which can never be met, so there is no measured risk profile and no live entry. That makes this a fundamentally interesting, technically unfired idea. The single fact that would move me to act: a decisive daily close above the 50-day EMA — $964.69 for Micron (about 9.2% above the $883.12 close) or $220.85 for Nvidia (about 4.3% above $211.81) — ideally paired with the next insider filing cycle showing selling turning to net buying.

Trade now: the contrarian memory thesis is coherent, but no entry is live

Wait — the business thesis here is interesting, but today's entry conditions are not met, and one of the required conditions is unsatisfiable as written. The idea argues (per Jensen Huang's public comments) that memory glut fears are sentiment rather than fundamentals, and the fundamentals do not contradict that: Micron's latest quarter (ended September 3, 2026) showed revenue of $54.2B, up 30.8% quarter over quarter, with net income of $37.7B and a 69.5% net margin. That supports the thesis as a business case. It is separate from whether a price-based entry is live right now — and it isn't. The attached rule set requires, among other things, the price to be below $0 — a condition that can never be true. As written, this setup can generate zero valid entries, so "waiting" in the strictest sense means waiting forever. Because of a price-data mismatch between raw and adjusted series, historical performance for both MU and NVDA was not measured, so no backtested evidence exists to weigh against this. The honest reading: this is a fundamentally-attractive, technically-unfired idea that needs a rebuilt trigger before any action. What "wait" means concretely: watch for momentum to confirm recovery on the daily chart. For MU (last close $883.12, observed 2026-10-24), the 50-day exponential moving average sits at $964.69 — roughly 9.2% above the price — and a daily close above it would be the first sign the drawdown is ending. NVDA ($211.81) is closer, about 4.3% below its $220.85 EMA. The downside that matters: MU's annualized volatility over the past two years is roughly 53% to 71% depending on estimator, with a maximum drawdown near 43%, so any dip-buying without a trigger carries real tail risk. Neither name has a confirmed upcoming earnings date in the supplied data.

The Customer Itself Is the Bull Argument — and the Latest Numbers Agree

Takeaway: the idea's core claim — that the memory selloff reflects glut fears rather than actual demand weakness — is supported by Micron's just-reported Q4 FY2026 results, Nvidia's stated appetite for memory, and supply-chain investment, though these numbers come with an important scope note: the strategy's historical performance was not measured because MU and NVDA raw and adjusted price series use different bases, so this is a fundamental analysis rather than a validated trade signal. ### Micron's newest quarter contradicts the glut narrative Micron's Q4 FY2026 (period ended 2026-09-03) shows revenue of $54.2B, up 30.8% quarter over quarter from $41.5B, with net income of $37.7B (a 69.5% net margin) and free cash flow of $32.9B, up 87.1% from the prior quarter. If a genuine supply surplus were emerging, you would expect pricing and margin compression; instead gross margin rose to 86.8% from 84.6%. Full-year FY2026 revenue of $133.2B versus $37.4B in FY2025 — a roughly 3.6x jump — implies memory demand, not supply, has been the binding constraint. That is the arithmetic behind the idea's 'sentiment, not fundamentals' framing. ### The buyer's own testimony and capital Per CNBC (2026-10-22), Nvidia CEO Jensen Huang says he sees no sign of a 2029 memory glut and that demand 'can always use more' memory, describing neocloud AI data-center demand as 'off the charts.' Nvidia's own filings show why the words carry weight: $96.2B of quarterly revenue growing 17.9% sequentially. Per Reuters (2026-10-22), Tesla and Nvidia invested $2.8B in Samsung Electro-Mechanics — the customer is putting capital into the supply chain, which reads as a bet that demand stays ahead of capacity, consistent with the thesis. ### Profitability quality that funds the cycle Both names sit at the top of their sector: Micron's gross margin ranks in the 95th percentile and operating margin (80.7%) and free cash flow ($32.9B) in the 99th percentile of Information Technology peers; Nvidia's free cash flow ($34.9B in its latest reported full year) also ranks in the 99th percentile. High margins plus strong cash generation mean memory makers can fund capacity expansion without balance-sheet stress, making a durable under-supply more likely to persist — and making the shares more resilient if demand holds. ### Both companies now return cash to shareholders Micron raised its annual dividend 22.8% to $0.46 per share for 2026 (trailing twelve months $0.565) and NVDA's annual dividend jumped from $0.04 to $0.51 per share in 2026. Dividends at this scale are only sustainable on the cash flows shown above, so the payout record is indirect confirmation that both managements believe current cash generation is durable — precisely the question the glut debate turns…

Scores

• Conviction score breakdown: 51
• Thesis support: 65
• Trade readiness: 20
• Risk quality: 42
• Fundamentals trend: 78

Watch items

• MU — Daily close vs 50-day EMA
• NVDA — Daily close vs 50-day EMA
• MU — Insider net open-market activity (provisional, period ending 2026-09-30)
• NVDA — Insider net open-market activity (provisional, period ending 2026-09-30)
• MU — RSI (14)