By Reshma Kapadia

Mention technology stocks, and investors almost instinctively think of the U.S. But Beeneet Kothari, founder of hedge fund Tekne Capital Management, believes the best opportunities in tech right now are overseas.

Kothari, who began his investing career as a technology analyst for the renowned Stan Druckenmiller, oversees $1.5 billion at Tekne, investing in a concentrated global portfolio of tech stocks that has beaten the MSCI World index by double digits for the past two years.

After foreign markets trounced the S&P 500 last year, some investors found a renewed appetite for diversification abroad. But Kothari was early to the trend. For instance, he saw the potential for China's artificial-intelligence development well before Chinese start-up DeepSeek reignited broader interest in Chinese stocks.

Kothari likes to go deep, often finding new ideas by interviewing customers of companies and following up on tidbits gleaned from industry executives. He looks for quality companies that others have underestimated. Barron's spoke with Kothari about which companies might be tomorrow's Taiwan Semiconductor Manufacturing or ASML Holding, and how China's AI buildout is different than in the U.S. An edited version of the conversation follows.

Barron's: Are the most attractive AI investments in the U.S. or China?

Beeneet Kothari: It is a no-brainer; China is a better investment proposition. China is going to do with AI what it has always done -- build things faster and cheaper. Sometimes the [state-driven] top-down mentality is the right approach. You need power, chips, data centers, and cooling. China is good at moving mountains to build things. It is a clear advantage.

When you make an investment decision today, the biggest thing you worry about is whether supply in the industry will double or triple next year. That is a legitimate risk in the U.S. because a lot of money has gone into the ground. China is emerging from a deep bear market. Because capital hasn't been invested in long-duration projects, you can have comfort in knowing you aren't going to have oversupply.

Also, the big AI companies in the U.S., including OpenAI, Anthropic, and Google DeepMind, are obsessed with artificial general intelligence and artificial superintelligence -- when is it happening, what does it mean, how much job displacement will it create? In China, which is in a recession, all a company can focus on is how to make money. It is an incredibly practical approach.

China's stock market was one of the strongest in the world last year, due in part to AI stocks. Where are valuations still attractive?

The most attractive valuations are in smaller and midsize businesses. We don't own Alibaba Group Holding or Tencent Holdings, which are more like Google.

We are one of the top five shareholders in GDS Holdings, a data-center business that is the landlord of AI. China is the largest country in the world, with the largest tech market. Data centers are the hottest thing in tech. Two publicly listed data-center companies in China -- GDS and VNET Group -- combined have a market value of about $12 billion. That is insane! You can't get the cheapest data-center stock in the U.S. for $10 billion.

People have discovered Taiwan Semiconductor and ASML. They are like the gateway drugs to international technology investing. When we bought Taiwan Semi, the stock was TW$500 a share [$15.90]. Now it trades for TW$2,000 and is among the top five holdings in many portfolios. There are several other layers in the tech supply chain, with scores of companies across Asia that are just as integral, albeit not as well-known.

Give us some examples.

As chips become more complex, we think the money will be made in advanced packaging. We own Tongfu Microelectronics, a $10 billion market-cap company, which has had a joint venture with Advanced Micro Devices since 2016.

China has passed a law that says 50% of all semiconductors produced in the country must use local supply chains. Tongfu is a winner, as is ACM Research. Both are part of the semiconductor capital-equipment supply chain in China.

China consumes 35% of the world's semis today and produces 6%. The government has said it wants to bring those numbers into alignment.

Tongfu trades for 13 times Ebitda [earnings before interest, taxes, depreciation, and amortization], with revenue growing more than 20% annually and structural growth behind it. A lot of these companies are underearning because they're young. Western peers are 40 years old and trade for 11 to 12 times Ebitda, with a fraction of the revenue quality and growth.

Investing in China, especially a strategic sector like semiconductors, comes with the risk that the U.S. might look to restrict such moves. How do you account for that?

It's always a concern, but the question is if that's priced in. Plus, the trend is also in the right direction. There's a thawing in tensions between the U.S. and China. Trump and Chinese leader Xi Jinping are going to meet four times this year, and officials have said the relationship with China is in a good place right now. We just allowed China to purchase Nvidia H200 chips.

Also, the export restrictions and moves over the past eight years, with the first Trump administration and then Biden, scared the Chinese and pushed them to localize and become more independent. That has made them a bit more resilient and highlighted how sensitive we are to [their dominance in] rare earth. It also created an opportunity in these semiconductor and supply-chain bets.

Baidu has been a big performer over the past year for the fund. What now?

Last summer, Baidu's stock was trading at cash. It is up more than 50%, spurred by plans for a spinoff of its AI chip unit, Kunlunxin. But nothing has happened yet. Once Kunlunxin's Hong Kong listing is executed, then will come proof of value for Baidu's Apollo robo-taxi business, which just hit 250,000 rides weekly -- on par with Waymo. Its agentic AI business, GenFlow, and CoreWeave-like HPC cloud will follow.

This is going to be the year of Chinese semiconductor initial public offerings. Some say Kunlunxin could be a $30 billion IPO. Alibaba is going to spin out its T-Head chip unit. CXMT, the largest memory company in China, is going to do an IPO this year. That could bring more investors. Plus, when semi companies earn or raise money, they spend it because they're always building new factories, new construction. There's going to be a lot of new capital raised, and we expect a massive new wave of capacity expansion to ensue.

More broadly, this could be the year for some of the big IPOs in China we've been waiting for -- fintech Ant Group, ByteDance, Shein, and Didi, the Uber of China, which we own. ByteDance is the biggest, most successful, and most global Chinese company, and most investors have no idea of it, but it's probably worth over a trillion dollars.

What is the biggest risk to the bullish case for China tech?

Their homegrown chips aren't quite ready. Access to Nvidia's even older-generation chips bridges that gap. If they lose that bridge, they are swimming naked a bit until 2028-30. Having this bridge probably accelerates the rate at which they can become self-sufficient.

What do you own outside of China?

Samsung Electronics' share price is off very low levels because it is the leading memory player in the world. But that isn't what interests us. Inside Samsung is the second-largest foundry after Taiwan Semi, and it has been sitting dormant.

Taiwan Semi took off because it landed Apple as a customer. Elon Musk and his conglomerate of companies -- SpaceX, Tesla, and xAI -- are now in need of a lot of chips and couldn't get much capacity out of Taiwan Semi, so he went to Samsung. Given the number of chips he needs over the next three to five years, Samsung could build out its foundry to rival Taiwan Semi.

Currently, Samsung is an $800 billion company, with most of its profits from memory. There is a business bigger than memory that no one is talking about, and it's about to get a shot in the arm with what could be the biggest IPO in history this summer with SpaceX, a company we know well through various relationships.

What is the significance of the offering?

The way foundries work is that your customer has to commit, and then you go out and spend. There were filings in the Korean stock market late last year that indicated that Tesla has committed to about $17 billion of chip purchases from Samsung. And Samsung is building out factories in Texas. What Apple did for Taiwan Semi is potentially what Elon could do for Samsung.

What are opportunities outside of Asia?

Brazil is the cheapest market in the world, with the index trading at about six times earnings. The central bank governor is going to change, and they are starting an easing cycle. Plus, there's a commodity boom.

We are also seeing a structural change in emerging markets. Companies in bear markets are finally behaving the way American companies have learned to behave: When you can't control the macroeconomics, you control capital allocation. We saw Chinese companies do this a couple years ago, when the economy was struggling and they became aggressive with buybacks.

Brazil is now doing that. In the next two years, fintech StoneCo, which we own, may return to shareholders over a third of its market cap through dividends or buybacks. The company is growing at about a 10% pace, above gross domestic product. But largely the story is that it's a concentrated industry and finally can be more efficient and grow profits.

Give us a contrarian pick.

(MORE TO FOLLOW) Dow Jones Newswires

February 12, 2026 01:30 ET (06:30 GMT)

Copyright (c) 2026 Dow Jones & Company, Inc.