Micron's ROIC finally exceeded its cost of capital at 15.68% for the first time in six years, yet insiders have sold stock in 15 of the last 20 quarters. The company that swung from -47% to 68% operating margins in seven quarters presents a paradox: is this sustainable value creation or just another peak before the inevitable memory cycle crash?
The current profitability metrics represent historical extremes that violate base rates for sustainability
Operating margins at 67.6% (98th percentile), up from -46.9% in Q2'23. ROIC at 15.68%, first time above WACC since Q3'18.
Management's capital allocation timing has destroyed shareholder value despite operational excellence
Buybacks executed at $1,086 average price versus current $338, resulting in -68.89% return on $5.4B spent.
The earnings yield to treasury spread is deeply negative, reflecting either cycle risk or growth expectations
2.9% earnings yield versus 4.3% treasuries creates 430bp negative spread at 8.5x P/E.
Is 85.5% revenue growth with a PEG of 0.10 a screaming buy or a cyclical trap?
Growth at this valuation compensates for any cycle risk
PEG of 0.10 with 85.5% TTM growth. ROIC exceeding WACC by 466bp. Market implies only 5% growth despite current expansion.
Peak cycle metrics masquerade as sustainable value
440% above DCF value. Operating margins swung 114 points in 7 quarters. Insiders selling during recovery.
Do insiders selling 15 of 20 quarters signal cycle wisdom or missed opportunity?
Insiders understand memory cycles and are taking profits at the peak
Net selling of $25M during profitability recovery. Only one buying quarter (Q4'25) in five years.
Institutions see value insiders are missing
Institutional ownership rising to 78.4%. 681 new positions versus 126 closed in Q4'25.
The 55-point spread between Lynch (0.75) and Marks (0.20) reflects genuine disagreement about whether this is a growth story at reasonable valuation or a cyclical peak. This wide divergence creates opportunity for investors who can correctly time the memory cycle.
None of the frameworks adequately weight Micron's technology leadership position in the AI memory hierarchy. While all focus on cyclical dynamics, they miss that HBM (high-bandwidth memory) commands structural premiums and multi-year customer commitments that could dampen traditional memory volatility. The company's segment concentration in DRAM (77.1%) positions it uniquely for datacenter AI workloads that require speed over storage.
If insiders who've navigated memory cycles for decades are selling at 68% operating margins while institutions accumulate at 8.5x earnings, who understands Micron's future better?