KOSPI Reclaims 5,800 Despite Intel Shock; October Exports Reach $115.3 Billion, Shipbuilders and Low-PBR Stocks in Focus

Despite a sharp drop in U.S. tech stocks, the KOSPI rose as much as 1.3% intraday to 5,807. The KOSDAQ also advanced more than 1.2%, reaching 733. Late in the session, the KOSPI climbed to 5,875 and the KOSDAQ to 736, extending their gains.

Although the Philadelphia Semiconductor Index fell more than 7%, the number of advancing stocks increased in the domestic market. More than 600 stocks rose on the Korea Exchange, while the KOSDAQ recorded around 180 gainers. Foreign investors were net sellers of about 200 billion won on the exchange, but institutions supported the index with net purchases of approximately 800 billion won.

Financial investment firms and pension funds participated in institutional buying, while purchases by other corporations also stood out. One explanation was that SK hynix’s share buyback was reflected in flows from other corporations. The won-dollar exchange rate rose by about 9.50 won to exceed 1,343 won, making it important to monitor foreign investor flows alongside exchange-rate movements.

This week brings the U.S. midterm elections, employment data and earnings announcements from major technology companies. Results are due from Apple, Palantir, AMD and Qualcomm, as well as the U.S. Job Openings and Labor Turnover Survey and the October Employment Situation report. In Korea, earnings announcements are concentrated among POSCO Group, EcoPro Group, and companies in the secondary-battery, cosmetics, gaming, telecommunications and entertainment sectors.

OpenAI CEO Sam Altman visited Korea on October 30 for a two-day, one-night trip. He met with Samsung Electronics and executives from Samsung Group, and had lunch with SK Group Chairman Chey Tae-won and executives. Semiconductor cooperation was discussed, but no specific supply contract was announced.

OpenAI’s fundraising needs were cited as a possible reason for the visit. According to estimates mentioned on the broadcast, OpenAI had initially sought to raise $30 billion but secured only about $11 billion, prompting discussion of lowering its target to around $20 billion. With large funding needs persisting, including amid Anthropic’s reported plans for an IPO, there was speculation that Altman may have discussed fundraising or ecosystem cooperation with Korean companies.

Some argued that if Korean semiconductor companies participate in OpenAI as equity investors, this could affect the company’s funding conditions and how it is valued in the corporate bond market. However, it has not been confirmed whether attracting investment was an actual purpose of the visit. Any follow-up announcements should distinguish between semiconductor supply cooperation and fundraising discussions.

Hanwha Ocean won a contract to build one U.S. Navy vessel at Philly Shipyard. The contract is worth approximately $269 million, with the possibility of building three additional vessels also open. Construction is scheduled to begin in the fourth quarter of next year, with delivery planned for 2030.

Expectations grew that U.S. naval contracts could turn the MASGA project into a tangible business opportunity, and Hanwha Ocean shares jumped more than 11% intraday. Given concerns about the aging U.S. fleet and its disadvantage relative to China, the U.S. shipbuilding and maintenance market could expand. Attention focused on Hanwha Ocean’s potential to secure business after acquiring Philly Shipyard.

However, Philly Shipyard recorded a net loss of about 120 billion won last year and was fully capital-impaired at the end of the year. The current contract remains limited in scale, so follow-on orders, the shipyard’s recovery and investment costs all warrant attention. Shipbuilding stocks more broadly have recently pulled back amid exchange-rate concerns, earnings worries and delays to projects tied to the U.S. political environment.

An analysis presented on the broadcast found that memory accounts for more than half of global cloud companies’ investment budgets, with the share of related spending reaching around 70% when GPUs are included. Rising memory prices have emerged as a pressure on Big Tech’s capital expenditure and free cash flow. It is important, however, to distinguish whether memory’s budget share has grown because prices rose or whether stronger demand drove prices higher.

It was also noted that increased investment spending does not necessarily mean data-center capacity itself is expanding. Even when building a data center of the same size, higher memory prices can increase the required budget. Investment totals alone therefore do not show that the number of data centers or AI computing capacity has grown by the same amount.

Goldman Sachs’ forecast was cited that free cash flow at Amazon, Google, Meta, Microsoft and Oracle would turn negative in 2026, widen to a deficit of about $150 billion in 2027, and then return to a surplus of $60 billion in 2028. By contrast, another analysis pointed to growing signs of AI monetization in cloud revenue. A key dividing line for the AI industry is whether investment costs will first convince the market or actual monetization will be demonstrated first.

Exports reached $115.3 billion in October, the second-highest total on record. Cumulative exports for the year reached $930 billion, raising the likelihood that the annual total will surpass $1 trillion in November. Semiconductor exports rose 259% year over year to $56.4 billion.

DRAM exports hit a record, while exports of MLCCs and PCBs also performed well. Non-semiconductor products, including large power transformers, instant noodles and cosmetics, also contributed to export growth. Semiconductors and SSDs accounted for 53.8% of total exports in October, down from 55.4% in September, while exports excluding semiconductors also showed a recovery from their August low.

Whether strong exports translate into improvements in household conditions and domestic demand remains a separate challenge. Retail sales returned to decline after five quarters, while durable-goods sales fell about 7% year over year on a quarterly basis and passenger-car sales dropped 11%. One explanation was that the base effect from household recovery coupons had faded and more individuals were cutting consumption after losses on stock investments.

Concerns were also raised that export growth was not being fully reflected in stock prices or domestic demand. One outlook held that a key factor for the 2027 economic outlook would be whether export gains spread to household income and consumption through wages and dividends. It was also noted that improvements in corporate earnings may take time to translate into performance bonuses and dividends.

The prevailing view presented was that Democrats were likely to take control of the House in the U.S. midterm elections. The Senate outcome could depend on results in closely contested states, making it difficult to call the race. If divided government becomes a reality, constraints on the administration’s budget expansion could improve policy predictability, though legislative gridlock and shutdown risks could also increase.

Opposition to data-center construction is being used by both Republicans and Democrats. In the United States, governors can influence data-center permitting, making gubernatorial races and state policies, including those in Texas, important alongside the federal elections. The impact of the election results on AI data-center construction and investment timelines warrants attention.

Reports said that 12 hours of talks related to Iran had taken place in the Middle East, and The Wall Street Journal reported developments concerning negotiations in the region. WTI held near $100 a barrel while falling about 3%. There were also reports that the United States had called for a swift resolution to the issue of Russia supplying gas to Europe through Ukraine.

Ahead of the U.S. midterm elections, speculation included both the possibility that President Trump would pursue negotiations to end the war and the prospect of further military action. Campaign rhetoric should be distinguished from actual progress in negotiations, while the effects of war and geopolitical tensions on oil prices and financial markets also warrant close attention.

The Korea Exchange published a list of low-PBR companies as part of efforts to address the Korea discount. The criteria were the bottom 25% of companies by three-year cumulative PBR within each KOSPI sector and the bottom 10% on the KOSDAQ. After excluding companies that submitted improvement plans, the list included 96 KOSPI companies and 55 KOSDAQ companies, for a total of 151.

The companies on the list had an average PBR of about 0.24 times. PBR compares a company’s net assets with its market capitalization; a low figure means the market values the company below its book net assets. The companies are expected to be flagged as low-PBR companies on HTS and MTS platforms.

The exchange and financial authorities plan to encourage companies to submit and implement improvement plans. One proposal discussed was to link monitoring of implementation to the stewardship code for institutional investors, including pension funds. Japan’s experience in improving undervalued companies was cited as a reference, including stronger disclosure and scrutiny of board decisions.

Lee Kwang-soo stressed that the purpose of keeping investment records is not merely to note the results of buying and selling, but to learn from the decision-making process. The idea is to record, at the time of purchase and sale respectively, the reason for buying, the stop-loss price, why events differed from expectations and the rationale for selling. He said a trading journal template would be shared as a PDF on the program’s page.

The trading journal for Leeno Industrial, presented as an example, listed its test-socket and pin-probe businesses and its inclusion in related ETFs as reasons for buying. After news of the chairman’s block sale sent the stock sharply lower, the investor closed the position and recorded both an attachment to the stock and the reasoning behind the sale. A producer’s Naver trade was presented as an example of relying on expectations around Jensen Huang’s visit and personal preference: after setting a 200,000-won stop-loss level, the producer sold at 216,000 won.

A producer who bought Kolmar Korea at 91,000 won sold at 87,600 won before reaching the stop-loss level out of anxiety; the stock later rose to 130,000 won. In the power-equipment ETF example, the investor continued holding based on the power bottleneck facing AI data centers and LS ELECTRIC’s orders. Lee said that using long-term business prospects and profit potential as the basis for a purchase, rather than short seasonal demand or other people’s opinions, makes it possible to correct the causes of mistakes.

Rather than remaining stuck watching an existing account with losses, Lee proposed starting the record-keeping process again in a new account by selecting index ETFs, semiconductor stocks or similar investments. His personal view was that the Korean stock market still offered worthwhile investment opportunities. The point was to build a new investment process and review it together rather than dwell on past mistakes and blame oneself.

Exports and the KOSPI have moved together over the long term, but recently they have diverged, with stock prices correcting even as exports rose. The fact that strong exports have not improved households’ day-to-day economic experience remains another concern. One view is that the next stage of export-led growth depends on whether corporate earnings gains spread through wages and dividends and revive domestic demand.

One interpretation was that the KOSPI had corrected after rising rapidly enough to rank among the world’s top-performing markets. This does not mean the correction was justified, but if the gap between export growth and stock prices narrows, economic conditions could provide a basis for investment. It was also argued that AI investment should be assessed not only in terms of short-term cash-flow pressure but alongside changes in cloud monetization and financing conditions.

The hosts stressed that buying and selling stocks based on attempts to predict the outcome of the U.S. midterm elections is risky. The point was to assess the long-term prospects of industries and companies when political shocks roil stock prices. They also argued that, given the United States’ reliance on South Korea’s shipbuilding, semiconductor, battery and nuclear power capabilities, Korean companies should be treated with respect on visa and cooperation issues.

Criticism was raised that publishing the low-PBR list should not end with publicly shaming companies. One argument was that improving shareholder value requires both incentives for companies that make improvements and penalties for those that fail to do so. The view was that the interests of ordinary shareholders should take priority over corporate inheritance and governance issues, and that the public should monitor whether the list actually shrinks.