Samsung: Twice the Share, and a Different ObjectiveThin moat

Samsung held about 29% of NAND revenue in the first quarter of 2026 against Kioxia's 14%¹. The share gap is the smaller half of the problem.

The larger half is that Samsung's flash business sits inside a company that also leads DRAM, competes in high-bandwidth memory, runs a foundry and sells the phones the memory goes into. That changes what the business is for. A pure play must earn its cost of capital in flash or it does not earn it at all; a division can be run for share, for capacity utilisation, or to keep a customer relationship warm for a product sold elsewhere.

In a commodity industry this is decisive at exactly the wrong moment. The participant with the deepest balance sheet and the least need to make money on this specific product sets the floor under prices, and everyone else discovers where the floor is.

What limits it is that Samsung is a rational operator with shareholders of its own, and it has not historically run flash at a loss for sport. It also arrived late to high-bandwidth memory, which is a reminder that scale does not guarantee execution.

The number to watch is not share but Samsung's flash capacity additions. A conglomerate expanding into a soft market is the specific event that ends a pure play's good year.