If Micron has only two years of decent forecasts, its limited visibility warrants a lower multiple, not a higher one. x.com
DarvasBoxAI
Momentum Trader, Darvas/TheStrat Systems, 30 Years’ Experience
Micron is too risky to own long term because memory is one of the most cyclical parts of the AI trade. Its margins are hard to forecast, so memory, like neoclouds, deserves a low multiple until it proves it can run a sustainable business through cycles, despite immense upside. Micron belongs in the too-hard pile: worth swing trading at opportune times, not holding long term or in real size. x.com x.com
Micron sits firmly in the too-hard pile because there is no clear view of what its earnings will look like three to five years from now. Memory stocks are simply not a fit here, with no criticism of memory investors. x.com
Micron looks quite cheap despite pulling back after an incredible earnings report, as one of only seven US companies over $10B with a forward P/E under 10 and 25%+ two-year revenue growth. Micron and Sandisk are cyclical, but each expects about 100% revenue CAGR over the next two years. They may not deserve 30-40x forward P/Es, but assigning a single-digit P/E is crazy. x.com x.com
Micron looks expensive on price-to-book at about 8x versus a historical average of about 2x, and relying on forward P/E for such a cyclical business can be dangerous. x.com
Micron's projections are set so high that even great years may not be enough to meet them. Samsung's Q3 results show the risk: revenue of $145.6B missed the $150.7B estimate, and operating profit of $80.2B missed $81.1B. x.com
Micron may merit 10x next year's earnings, perhaps a $2,000 share-price target rather than D.A. Davidson's $3,000, on the firm's case that AI and memory demand could support three to five years of growth. x.com
Micron faces a severe crash risk from CXMT's current threat and apparent investor complacency after a euphoric high. If the Wall St Cheat Sheet fits, most of Micron's gains across 40 years, amassed in just one year, could be erased soon. Earlier, Elon Musk said Micron had been reasonable on future memory prices; memory stocks were not yet to be faded. x.com x.com
Semiconductors could stay undersupplied for a long time as more capable AI per dollar creates more uses and drives compute demand. Micron and NVIDIA trade at low forward P/Es because investors fear peak cyclical earnings, yet current earnings show no slowdown. Shortages may worsen as AI adoption spreads, though the industry could remain cyclical; a reversal of the forced rollover trade is hoped for, not assured. x.com x.com x.com
Micron deserves 10-12x forward earnings rather than 6.2x, and its latest quarter only strengthened the bull case. Revenue reached $54.23B, up 379% year over year, and EPS $33.42, up 1,003%; next-quarter guidance is $62B revenue and $38.15 EPS versus estimates of $58B and $35.92. The market fears growth and demand could suddenly stop, but no one can time when or if that happens, whether in 1, 2, 3 or 5 years. It might not deserve 20x. x.com
Buyers rushing into Micron's gap-down open show fear is still too limited for a sustainable low. They bought hand over fist, while the equity put/call ratio stood at 0.75; significantly more pain and fear are needed. x.com x.com