A high-stakes confrontation is unfolding in South Korea as semiconductor giants Samsung Electronics and SK Hynix face mounting pressure from the investment community. The catalyst is the unprecedented liquidity generated by the demand for High Bandwidth Memory (HBM) chips, the backbone of the Artificial Intelligence revolution.

A Cash Mountain Raising Doubts

According to LSEG data, the two companies are projected to hold a combined $263 billion in net cash by year-end. This figure is staggering, more than double Nvidia's estimated $102 billion and exceeding the combined cash reserves of the remaining six companies in the U.S. "Magnificent Seven" group.

Despite record-breaking profits, the stock prices of both companies have retreated significantly from their June highs—SK Hynix by 48% and Samsung by 37%. Investors view the management's reluctance to commit to aggressive shareholder reward programs as a sign of doubt regarding the long-term sustainability of the AI cycle.

The "Korea Discount" and Shareholder Demands

Pressure is mounting from two sides. Institutional investors, including Janus Henderson Investors, argue that holding such vast reserves creates an inefficient balance sheet. Simultaneously, the ACT platform, representing retail investors, is calling for Samsung to initiate a $32 billion share buyback program. Key demands include:

  • Increasing the return of free cash flow to shareholders from 50% to 80%-100%.
  • A clearer capital allocation strategy to combat the so-called "Korea Discount."
  • Alignment with competitors like Micron, which recently pledged a 100% return of free cash flow.
"When you say you can't commit because the future is uncertain, you're just feeding the narrative that the AI boom is transient and cyclical," notes Richard Clode of Janus Henderson.

The Corporate Defense

Management at the South Korean conglomerates counters that the semiconductor industry requires massive capital expenditures (CapEx) to maintain technological leadership, with planned investments totaling approximately $2.07 trillion. However, analysts point out that multi-year supply contracts with major AI clients mitigate risks, suggesting that funding new production lines and boosting dividends are not mutually exclusive goals.