I am currently long MU, but might soon close that position by this year end. MU is by no means a monopoly as many newish MU bulls believe. It is still in a race-to-the-bottom commodity business. I made a ton and lost a ton on this stock. Be careful out there. Key points to ponder about:
1) MU's 2/3rd revenue and 80% of profits are from DRAM.
2) MU is atleast a half to one full node generation behind the industry leader Samsung. Everytime Samsung ramps up on the newer smaller node, oversupply ensues and customers gravitate towards newer and better product from Samsung. As DRAM is highly price-inelastic (for example, adding more DRAM than required doesn't improve performance in smartphones), the leader, atleast temporarily captures higher market share, only to give it away later as laggards like MU and Hynix catch up. This is what you see as the infamous Memory boom/bust cycle.
3) A counter to point 2 above though is the 'end of the runway' or the fundamental limits reached in what can be achived with node shrinkage in DRAM. When no further improvements can be made in DRAM manufacturing technology, leaders and laggards will become one and the same, and that could herald a new era of oligopoly. Only then, boom-bust cycles will be a thing of the past and all three DRAM players - Samsung, MU and Hynix - will play nicely with each other to maximize their profits. I am expecting one more boom-bust cycle before they reach oligopoly nirvana.