i get why most banks have higher price targets on $MU
their thesis varies but mine is very straightforward.
micron kept 84.6 cents of every dollar it billed last quarter. a year earlier it kept 37.7.
this market is heavily underestimating micron's future cash generation potential.
revenue went $9.3bn to $41.5bn over that year, so gross profit went up ten times.
-- that markup is the entire reason kioxia is putting upgraded NAND behind CXL to displace some DRAM.
-- and the reason CXMT has a 3D DRAM roadmap at all. risk production 2028, mass production 2029.
-- every one of those workarounds lands after the contracts being signed this year have repriced.
nobody engineers around a normal margin.
real question is what they do with the cash once the CHIPS act restriction lifts in december:
-- return to shareholders (they’ve said 100% of excess cash, buybacks first) or
-- keep pouring it into capacity so the 85 cent on the dollar keeps working beyond the contracts already getting signed
-- or go for a bigger ecosystem play (more like what
$NVDA did with its frontier lab + neocloud strategy)
their thesis varies but mine is very straightforward.
micron kept 84.6 cents of every dollar it billed last quarter. a year earlier it kept 37.7.
this market is heavily underestimating micron's future cash generation potential.
revenue went $9.3bn to $41.5bn over that year, so gross profit went up ten times.
-- that markup is the entire reason kioxia is putting upgraded NAND behind CXL to displace some DRAM.
-- and the reason CXMT has a 3D DRAM roadmap at all. risk production 2028, mass production 2029.
-- every one of those workarounds lands after the contracts being signed this year have repriced.
nobody engineers around a normal margin.
real question is what they do with the cash once the CHIPS act restriction lifts in december:
-- return to shareholders (they’ve said 100% of excess cash, buybacks first) or
-- keep pouring it into capacity so the 85 cent on the dollar keeps working beyond the contracts already getting signed
-- or go for a bigger ecosystem play (more like what
$NVDA did with its frontier lab + neocloud strategy)
