InterviewAI-fueled rally at Samsung, SK hynix faces sustainability test: Harvard economist

John Y. Campbell, a professor of economics at Harvard University, speaks during an exclusive interview with The Korea Times and its sister publication, the Hankook Ilbo, at the Korea Capital Market Institute in Seoul, Wednesday. Korea Times photo by Yun Gi-hun

Samsung Electronics and SK hynix have led Korea’s recent stock market rally on the back of surging global demand for advanced artificial intelligence (AI) chips, but the broader AI industry must ultimately prove it can translate technological promise into scalable and sustainable profit models for current lofty valuations to hold over the long term, according to a Harvard economist.

In a recent exclusive interview with The Korea Times and its sister publication, the Hankook Ilbo, at the Korea Capital Market Institute in Seoul, John Y. Campbell, a professor of economics at Harvard University, noted that earlier waves of the tech economy justified massive valuations because dominant platforms — Google in search, Amazon in e-commerce and Meta’s social media empire — captured highly lucrative markets through powerful network effects, effectively securing near-monopolistic positions.

The AI industry, however, may not follow the same trajectory, as relatively low switching costs and the rapid emergence of competing models raise uncertainty over whether sustainable market dominance can be established.

“We have competing models and it seems to be very easy to switch among them,” Campbell said, noting that AI firms may struggle to replicate the entrenched market control that once underpinned tech giants.

Unlike AI model developers whose valuations hinge heavily on future monetization, the current semiconductor rally pushing shares of the two Korean chip giants to record highs is grounded in more immediate, tangible conditions — hyperscalers’ soaring need for high-bandwidth memory (HBM) and aggressive data center expansion.

Geopolitical tensions are further strengthening Korea’s semiconductor competitiveness, Campbell noted, as countries and companies are increasingly seeking alternative chip supply sources outside Taiwan for strategic reasons.

“The Korean companies are in a good position to exploit that,” he said.

Still, there are growing warnings that valuations could come under pressure if investors become overly optimistic about the long-term growth potential. If major tech firms pull back on AI infrastructure spending, today’s breakneck expansion of chip production could trigger oversupply. At the same time, rising output from Chinese manufacturers may intensify competition, particularly in lower-end memory segments.

Campbell underscored that ultimately, the long-term performance of Korean tech stocks will depend not simply on AI-driven hype, but on whether companies can sustain their technological edge and deliver consistent earnings growth.

Drawing on his long-standing research into market valuations, the Harvard professor also cautioned that elevated stock prices themselves may point to more muted future returns.

“Historically, high prices relative to current earnings tend to go along with low subsequent returns,” he said, citing his influential work on the cyclically adjusted price-to-earnings ratio, or CAPE. “I’m not saying there’s going to be a crash, but at today’s prices, you should expect lower returns than you would have a year ago.”

Only AI firms capable of leveraging their models to create genuinely transformative, patent-protected products that can “invade some other industry” may justify today’s premium valuations, he argued.

A display board at the Korea Exchange in Seoul shows the closing prices of Samsung Electronics and SK hynix, Wednesday, as the benchmark KOSPI closed above the 7,000 mark for the first time on the back of a semiconductor rally. Yonhap

Beyond his skepticism toward inflated AI valuations, Campbell’s latest work also turns to a more personal but equally consequential question — why modern financial systems so often work against the very people who need them most.

In his 2025 book, “Fixed,” co-authored with Tarun Ramadorai, professor of financial economics at Imperial College London, Campbell argues that personal finance markets frequently disadvantage ordinary consumers because many financial products are unnecessarily complex, behaviorally exploitative and structured to favor powerful institutions.

"Fixed" by John Y. Campbell and Tarun Ramadorai / Courtesy of Princeton University Press

The issue has drawn growing attention in Korea as well. On May 2, Kim Yong-bum, presidential chief of staff for policy, questioned the fairness of a credit system in which “financially secure borrowers enjoy the lowest rates while those in the most precarious situations face the highest borrowing costs.”

Campbell acknowledged that volatile incomes and higher default risks naturally make lenders more cautious toward lower-income borrowers. But he stressed that the system’s inequities run deeper than simple credit assessments.

“It’s not just about the credit risk,” he said. “It’s also that the system creates products that are too complicated and hard for people to use, and it doesn’t encourage people to build emergency funds or to arrange credit in advance.”

Instead, he advocates for simpler, standardized financial “starter kit” products that are broadly accessible, easier to navigate and less vulnerable to hidden fees or manipulative bundling.

He pointed to Korea’s “jeonse” system — the country’s unique lump-sum rental deposit model — as an example of how two completely different products, rent and credit, are problematically bundled together.

Under the system, tenants effectively provide landlords with large upfront loans through refundable deposits, exposing renters to financial risks that many are ill-equipped to evaluate. “Ordinary people who want to rent an apartment are not in the business of evaluating credit risks,” he said.

While AI is often touted as a tool that could improve consumer financial decision-making, Campbell remains cautious.

“The question is whether we can trust AI agents to give us good advice,” he said, warning that many systems may be trained on existing financial information already shaped by incumbent institutions whose interests do not always align with consumers.

“There’s a real issue about who the AI is working for,” he added, stressing the need to question whether AI is truly providing the best advice to consumers or merely reinforcing conventional wisdom shaped by the interests of financial institutions.