AI chip demand keeps piling up — ride the momentum in Micron

Verdict: the AI memory story is real, but the trade isn't live yet — wait for confirmation

The demand story is genuinely in the numbers. Micron's fiscal Q3 (ended May 28, 2026) showed revenue of $41.5B, up 73.8% quarter over quarter, gross margin at 84.6%, and free cash flow of $17.6B, up 218.4% — while the Tencent deal to lease 100,000 chips through Oracle (per the Reuters report on the FT story published October 1, 2026) adds a second demand source. The strongest case against is uncomfortable on both legs. The June 30, 2026 ownership filings show net open-market insider selling of roughly $231.1M at Micron across 29 holders and $45.0M at Oracle across 26 holders. Micron's own history — gross margin swinging from negative 32.7% in 2023 to today's 84.6% — marks these as cycle-peak economics, and Oracle burned $5.4B in free cash flow last quarter. Meanwhile, the trade itself is not live. ORCL at $137.30 still sits below its $143.85 trend average, with RSI at 38.7 needing to cross 45 and ADX at 13.0 needing to top 20, and the MU leg has already run far past its dip-buy level — so there is nothing to execute today. The 60-month completed backtest shows a 12.4% return across 7 trades with a 71.4% win rate but a 33.7% maximum drawdown, and no robust parameter setup was established because the final holdout produced no evaluable trades. Conviction breakdown: thesis support 72, fundamentals trend 78, trade readiness 35, risk quality 42, backtest evidence 48.

Trade now: no entry yet — both legs need a pullback-and-reclaim, not chasing strength

The short answer today is: wait. The MU leg is the furthest from firing. At $1,065.11, Micron has already reclaimed its 21-day trend average ($1,017.98, now $47.13 above the entry mark) and sits above the Bollinger midpoint ($1,012.67), while RSI (14) at 61.0 is well past the 45 cross level and ADX (14) at 17.5 still trails its 20 threshold. In plain terms, the strategy wants to buy a dip into the trend average followed by a momentum reclaim — not to chase a price that has already run 486.5% off its range low. Until MU cools back toward the low $1,010s and rebuilds momentum from there, there is nothing to execute. The ORCL leg is closer but still incomplete. Oracle closed at $137.30, below both the 21-day average ($143.85, only $6.55 away) and the Bollinger midpoint ($147.11 — that condition is already met). What's missing is momentum: RSI (14) at 38.7 must climb and cross above 45, and ADX (14) at 13.0 must push above 20. That means the wait here is days to weeks of trend-strength building, not a distant price level. If Oracle reclaims $143.85 while RSI crosses 45 and ADX tops 20 on the same signal, the entry triggers. On execution mechanics: the strategy's fixed-risk sizing uses a stop 2.4% below entry and a first profit target 4.9% above entry — roughly 2:1 reward to risk — with a secondary stop below the second-ranked support and profit at the 127.2% extension level. Over the supplied 60-month completed test, the pair strategy produced 7 trades with a 71.4% win rate, a 12.4% total return, and a worst peak-to-trough drawdown of 33.7%. That tells you the edge comes from few, selective entries — so missing the first few dollars of a move is an acceptable cost of discipline. Concretely, "wait" means: set alerts at $1,017.98 on MU and $143.85 on ORCL, and check daily whether RSI resets toward the mid-40s (MU) or crosses 45 (ORCL) with ADX rising through 20. One positioning note: both legs share the same demand thesis (per the idea's own framing, AI data-center memory demand plus Tencent leasing 100,000 chips through Oracle), and the two tickers show a 0.35 correlation over the past 500 sessions. The 25% maximum position size and 2.4% fixed risk per entry are what keep a correlated thesis from becoming a correlated loss.

Two demand engines and a confirmed earnings beat — the bull case has receipts

The demand story in this idea is not just narrative — it shows up in Micron's own filings. For the fiscal quarter ended May 28, 2026, revenue came in at $41.5B, up 73.8% from $23.9B the prior quarter, while net income more than doubled to $28.2B and free cash flow jumped 218.4% to $17.6B. Gross margin rose to 84.6% from 74.4% quarter over quarter, and return on equity climbed to 28.0% from 19.0%. Whatever you think of the trading setup, the fundamental acceleration the idea cites as confirmation is visibly in the numbers. The second, independent leg — Chinese AI players sourcing compute despite export limits — also has a factual anchor. Per the Reuters report on the Financial Times story published October 1, 2026, Tencent agreed to lease 100,000 chips from Oracle to accelerate its AI push. Oracle's most recent quarter (ended August 31, 2026) shows revenue of $19.3B, up 0.8% sequentially, with operating margin improving to 34.8% and net income of $4.8B — the kind of infrastructure-driven demand the idea describes. The Bloomberg segment the same day also framed the AI boom as powering Micron's earnings. On the quant side, the realized evidence supports the idea's claim that momentum after confirmed winners has tended to work here. The backtested MU pair over the 60-month window produced a 12.4% return across 7 trades with a 71.4% win rate. That is not a dramatic edge, but it is a positive, tradeable one — with the caveat that the equity curve…

Scores

• Conviction score breakdown: 55
• Thesis support: 72
• Trade readiness: 35
• Risk quality: 42
• Backtest evidence: 48
• Fundamentals trend: 78

Watch items

• MU — Close vs 21-day EMA
• MU — RSI (14)
• MU — ADX (14)
• ORCL — Close vs 21-day EMA
• ORCL — RSI (14)
• ORCL — ADX (14)
• MU — Insider net open-market activity (next 13F/Form 4 cycle)
• ORCL — Free cash flow (next quarterly report, fiscal Q2 ~early December 2026)
• MU — Fiscal Q4 results (quarter ending ~late August 2026)