Samsung's Record Profit Masks Foreign Investor Exodus

The Number That Doesn’t Tell the Whole Story

Samsung Electronics reported a provisional third-quarter operating profit of 107 trillion won ($72 billion). It is the highest quarterly figure in the company’s history, a milestone that underscored the sheer force of the memory‑semiconductor supercycle now driving global AI infrastructure buildouts. Memory prices for both DRAM and NAND have climbed steeply over the past two years, fueled by an unrelenting surge in demand from AI server farms, cloud hyperscalers, and enterprise data centers upgrading their storage architecture. Analysts at major Korean investment firms have responded by revising their full‑year earnings forecasts upward across the board, with several estimating that Samsung could end 2025 with an operating profit well above the 150 trillion won mark—a figure that would eclipse the previous annual record set only two years ago.

But the same quarter saw Samsung’s foreign ownership stake drop below 40 percent, a level market participants describe as crisis‑territory for a company of its stature. Over the past five years, foreign investors have sold roughly a third of their holdings, unwinding positions at a pace that has rarely been seen in Korean equity markets. The paradox is stark: record profits coexist with a sustained selloff that shows no sign of abating. Understanding why requires looking past the cyclical upswing in memory prices and into deeper structural anxieties about Samsung’s competitive position in the very technologies that are powering this cycle—notably high‑bandwidth memory and advanced foundry fabrication.

What the Analysts Are Saying

The securities research community is uniformly bullish on near‑term earnings, but the tone of their commentary reveals a subtle shift in what investors are valuing. Ryu Young‑ho of NH Investment & Securities noted that the diffusion of AI agents is expanding memory demand at both national and corporate levels, adding, “This time, the numbers will ultimately prove that it is different from the past.” He also suggested that while the broader market believes further price hikes are unlikely in the near term, the current demand levels could still support incremental increases through 2026, particularly in the enterprise and data‑center segments where supply remains constrained.

Kim Un‑ho at IBK Investment & Securities agreed that memory profits have exceeded expectations, estimating that price increases are outpacing market forecasts by a meaningful margin. He pointed to anecdotal evidence from Samsung’s largest customers—major cloud providers who have reportedly locked in multi‑year supply agreements at premium pricing, effectively insulating Samsung from the sharpest troughs of the traditional memory cycle. Park Yu‑ak of Kiwoom Securities observed a shift in market narrative: short‑term stock momentum has moved from “earnings‑upgrade flows” to “the direction and scale of shareholder returns.” In other words, analysts are no longer asking simply whether Samsung will earn more; they are asking whether Samsung will give more back.

That shift points directly to the next question: how will Samsung deploy its windfall? Baek Kil‑hyun of Yuanta Securities highlighted that the market is now pricing in mid‑to‑long‑term capital allocation, with 2027–2028 free‑cash‑flow expansion serving as a key benchmark. “It will be important how concrete management’s medium‑to‑long‑term capital‑allocation plan, including treasury‑stock purchases and cancellation, becomes,” he said. Son In‑jun of Eugene Investment & Securities added that an additional shareholder‑return announcement could come as soon as the fourth‑quarter results in January, potentially including a special dividend or an accelerated buyback program that exceeds the 10 trillion won already authorized for 2025.

The Shareholder-Return Dilemma

Samsung’s board faces a genuinely difficult balancing act. On one hand, the company carries one of the most ambitious capital‑expenditure programs in the semiconductor industry, with plans to invest more than 300 trillion won over the next five years in new fabrication facilities in Texas and Taylor, expanded R&D on next‑generation memory architectures, and the construction of a dedicated advanced logic foundry complex. These investments are essential to maintaining competitiveness, but they also represent a massive drag on free cash flow and limit the room for generous shareholder distributions.

On the other hand, the pressure from investors—especially domestic institutional investors who have grown frustrated with what they see as passive management—is mounting. The Korea Exchange’s recent push for improved corporate governance and the Securities Depository’s public discussion of a mandatory minimum dividend ratio for large listed companies have added further urgency. Samsung’s current payout ratio of roughly 15 percent of earnings lags far behind peers like SK Hynix, which has already signaled a commitment to return a significantly larger share of profits to shareholders.

If Samsung follows through on expanded buybacks and dividends, domestic retail investors and index funds that have stuck with the stock will benefit most. The share price would likely find a floor, and the yield could become an attractive substitute for capital gains in a low‑growth environment where Korean equities have underperformed regional peers for over a decade. Pension funds and insurance companies, which hold Samsung as a core portfolio asset, would also welcome a more predictable income stream.

Who Wins, Who Loses

Foreign investors, however, remain skeptical. For many global fund managers, the math is straightforward: Samsung’s profit cycle is tied to memory‑price volatility, and the company’s heavy reliance on cyclicality makes it a risky long‑term hold in portfolios increasingly dominated by secular growth themes. More importantly, they are watching the high‑bandwidth memory race—and in that contest, Samsung is trailing its domestic rival SK Hynix. While Samsung is still qualifying its HBM3E chips for NVIDIA’s next‑generation GPUs, SK Hynix has already secured the bulk of early orders from the world’s largest AI chip designers. The perception is that Samsung’s foundry and advanced‑memory segments are losing ground to both Chinese competition in legacy nodes and faster execution by rivals in cutting‑edge processes.

This competitive gap has created a second‑order effect: even as Samsung’s overall profit soars, the market is assigning a lower multiple to those earnings because it views a significant portion of them as non‑recurring. The “Korea discount” is therefore not just about governance; it is increasingly about growth trajectory. Investors are willing to pay a premium for companies with durable competitive advantages, and Samsung’s memory business—while undeniably profitable today—does not inspire the same level of confidence in the longevity of its market position as, say, TSMC’s foundry dominance or SK Hynix’s early lead in HBM.

The Real Disconnect

The “Korea discount” is not a new concept, but it has intensified in recent years. It reflects a bundle of structural concerns: corporate governance weaknesses, chaebol succession opacity, and a perception that Korean conglomerates prioritize market share over shareholder returns. Samsung’s 107‑trillion‑won profit would have been enough to silence critics a decade ago. Today, investors are asking whether the same strategy that built the empire—scale, vertical integration, and aggressive capacity expansion—can win in an era of AI‑driven memory demand where differentiation is measured in nanometers and bandwidth, not just gigabytes.

Memory‑price cycles have always been brutal. The last downturn in 2018 wiped out billions in market value and left Samsung with operating losses that shook investor confidence for years. When prices normalize—and they will—Samsung’s profit base will compress again, perhaps dramatically. Foreign investors are selling into the strength because they do not trust the sustainability of the current cycle. They are also worried that Samsung’s capital expenditure plans, which include a new Texas fab and heavy R&D spending on next‑generation memory, will limit the cash available for dividends and buybacks. Every won invested in a new fabrication plant is a won not returned to shareholders.

There is also a geopolitical dimension. The CHIPS Act in the United States has provided Samsung with substantial subsidies for its Texas operations, but it has also imposed strict conditions—including restrictions on expanding advanced capacity in China—that complicate the company’s global strategy. Chinese competitors, meanwhile, are receiving massive state support and closing the technology gap in legacy memory segments, putting pressure on Samsung’s lower‑margin businesses even as the company rides the high‑end AI wave.

What Comes Next

The critical catalyst is Samsung’s upcoming earnings call and the potential announcement of a concrete shareholder‑return plan. If management commits to a multi‑year buyback program of tens of trillions of won, combined with a clear timeline for HBM qualification and foundry recovery, foreign funds may pause their selling. A credible plan could also trigger a re‑rating of the stock as analysts upgrade their fair‑value estimates.

But if the plan is vague or modest—if Samsung merely continues its existing buyback rhythm without a significant increase—the exodus is likely to continue, and the Korea discount could deepen further. Domestic investors should watch the quarterly dividend per share and any special‑dividend signals closely. The stock’s valuation remains depressed relative to peers like SK Hynix, suggesting room for multiple expansion if capital allocation improves.

For foreign investors, the decision hinges on whether Samsung can prove it can compete in the next generation of memory technology while delivering consistent returns. The memory supercycle may last longer than anyone expects, but supercycles always end. What matters for Samsung’s valuation in the long run is not how much profit it generates this quarter, but whether it can build a moat around its leading‑edge products that survives the next downturn.

The Bottom Line

Samsung’s record profit is real, but it is also a reminder of how cyclical the business remains. The foreign selloff is not a bet against the current cycle; it is a bet against the next one. Unless Samsung transforms its capital‑allocation strategy, closes the HBM gap with SK Hynix, and demonstrates that its growth drivers extend beyond memory‑price volatility, the “Korea discount” will persist. The market will reward execution, not just earnings headlines—and Samsung’s next chapter will be written not in won, but in trust.