Samsung Electronics is reportedly planning to return more than 100 trillion won, equivalent to approximately $71.75 billion, to investors through a new payout program. The strategy is set to be presented for approval at a board meeting at the end of August and includes a special dividend funded by half of the firm's available cash.
The scale of the planned return reflects confidence in the sustainability of current earnings driven by the AI infrastructure boom. Memory is vital for AI servers, and returning capital rather than accumulating large cash reserves addresses previous investor criticisms. Samsung and SK Hynix had previously faced pressure for being conservative with shareholder returns despite recording massive profits, with analysts noting that holding excessive cash could signal uncertainty regarding the future of the AI sector.
Samsung's payout follows an announcement by rival SK Hynix regarding a share buyback and cancellation program totaling 40 trillion won, or about $28.67 billion, which represents the largest shareholder return ever announced by a publicly listed South Korean company. Together, the announcements indicate that capital generated during the AI hardware cycle is beginning to return to investors.
The funding for Samsung's program stems from an AI memory cycle that bolstered revenues. For the quarter ending June 30, the company reported 171.5 trillion won in revenue and an operating profit of 89.5 trillion won. The Device Solutions unit, responsible for memory production, generated nearly all of the profit due to high demand for high-bandwidth memory, server DRAM, and enterprise SSDs, while the mobile business experienced a minor operating loss.
Strong profitability among key memory manufacturers has coincided with sharp increases in memory prices. TrendForce projected that DRAM contract prices would rise between 58% and 63% in the third quarter, with NAND flash prices climbing 70% to 75%. Memory supply is expected to remain structurally tight through 2027 as production prioritizes higher-value AI components, a trade-off that may ultimately lead to higher costs for consumer devices such as personal computers and smartphones.
