Deep Dive — MU: memory boom tests near-high IV

Micron’s July 9, 2026 setup is not a quiet semiconductor story. The company’s fiscal third quarter, ended May 28, 2026, put a much larger memory cycle in front of option sellers: revenue of $41.46 billion, non-GAAP earnings of $25.11 per diluted share, and operating cash flow of $25.39 billion. That is enough business momentum to explain why the options tape is rich, but it is also the reason the assignment test needs to be stricter than a simple premium scan.

The scale of the step-up is the first fact that matters. Micron reported fiscal Q3 revenue of $41.46 billion, up from $23.86 billion in fiscal Q2 2026 and $9.30 billion in fiscal Q3 2025. On those rounded reported figures, revenue rose about 73.8% sequentially and about 345.8% year over year. The company also guided fiscal Q4 2026 revenue to $50.0 billion ± $1.0 billion, which keeps the next checkpoint anchored in the income statement rather than only in the stock chart.

The mix also matters. The Cloud Memory business unit reported $13.769 billion of fiscal Q3 revenue and Core Data Center reported $11.524 billion. Together, those two data-center-linked units represented roughly 61.0% of Micron’s reported fiscal Q3 revenue. Mobile and Client added $11.521 billion, while Automotive and Embedded contributed $4.634 billion. For a wheel trader, that mix frames MU as a high-beta AI and memory-cycle exposure, not as a slow industrial compounder.

The product notes explain why the market is treating the cycle as an AI infrastructure story. Micron said HBM4, built on 1-beta DRAM technology, was in high-volume shipments for its lead customer’s platform, and that HBM4E development on 1-gamma DRAM technology was underway with volume production expected in calendar 2027. Those details do not remove cyclicality, but they make the next few quarters unusually tied to data-center demand and customer qualification timing.

The options market is not ignoring that risk. Cboe’s delayed July 9 dataset showed MU IV30 at 99.20%, against an annual IV30 low of 38.37% and an annual high of 107.70%. In plain terms, the implied-volatility reading was about 88% of the way from its annual low to its annual high. That can make premiums look attractive, but it also means the market is explicitly charging for a wide range of outcomes.

A useful way to read the chain is to separate liquidity from valuation. Liquidity answers whether a contract can be quoted and exited with reasonable discipline; valuation asks whether the credit compensates for the business risk. MU clears the first screen better than many single names because the Cboe chain shows thousands of series, multiple expirations, large open interest and active volume. The second screen is harder. Near-high IV is a market warning label as much as a yield source, because large expected moves can erase the comfort created by a large initial credit.

The balance-sheet and product story also changes the covered-call conversation. When a company moves from a weak memory environment to a quarter with $41.46 billion of revenue and a $50.0 billion midpoint guide, a mechanical call-sale habit can cap participation in the very repricing that created the premium. That does not make calls unattractive by default; it means the strike and expiration need to reflect whether the shares are being held as inventory for income or as a business the holder still wants after a major reset.

• For covered calls: compare the call credit with the willingness to lose shares during an AI-memory repricing cycle.
• For cash-secured puts: treat the strike as a possible ownership price, not only as distance from the tape.
• For expiration selection: separate event risk from ordinary theta decay, especially when IV30 is near the top of its annual range.
• For position sizing: respect that a cyclical semiconductor can move faster than a diversified ETF when guidance changes.

The covered-call angle is therefore two-sided. A shareholder may see unusually rich call premium because the business is producing a large AI-memory revenue surge and the chain is liquid. The trade-off is that a high-IV call can still be cheap if the next leg of the cycle keeps repricing the stock faster than the option premium collected. Strike selection becomes less about the biggest credit and more about whether assignment would be acceptable after a fast fundamental reset.

The cash-secured-put angle is just as conditional. A put seller is not only underwriting a ticker; the seller is underwriting a cyclical memory business at a moment when gross margins, data-center demand, HBM supply, customer concentration, and export or capex headlines can all matter. The premium can be real, but so is the gap risk around guidance, AI infrastructure spending, and any sign that memory pricing is peaking.

The next checkpoints are specific. Micron’s fiscal Q4 revenue guide of $50.0 billion ± $1.0 billion sets the operating bar. Cboe’s IV30 range shows how much risk is already embedded in the chain. The business-unit table shows where the AI-memory thesis is concentrated. For July 9, 2026, that combination makes MU useful as a premium-seller case study: strong fundamentals, deep options activity, and an implied-volatility level that demands respect rather than autopilot.