This Fund Manager Bought Nvidia and SK Hynix and Sold Software Before Others. His Simple Message on AI: 'follow the Money.'

Dow Jones07-02 18:52

Stephen Yiu lines his fund with recipients of AI investments

Invest in the companies that are receiving artificial intelligence investments, says Blue Whale Growth Fund chief investment officer Stephen Yiu.

Investing in the artificial intelligence trade takes nerves of steel these days, with visibility ever-cloudy over whether all the money being invested will ultimately pay off.

A far simpler way to approach that complicated trade is to just "follow the money," says Stephen Yiu, the lead fund manager of the nearly $4.5 billion Blue Whale Growth Fund.

"Where's the money going? That's the question you have to ask. And then you try to position into companies that are going to receive that spending," Yiu, also chief investment officer of Blue Whale Capital, told MarketWatch in an interview on Wednesday. "For the AI spenders, what they're trying to do is spend the money first, hoping that they will make the money back later if they're successful in AI, but that's not a given."

Yiu's London-based fund is packed with global semiconductor industry players in the receiving line of those investments - Applied Materials $(AMAT)$ and Dutch-based BE Semiconductor (NL:BESI) at the No. 1 and No. 2 spots, with SK Hynix (KR:000660), Lam Research $(LRCX)$, Lumentum $(LITE)$, Nvidia (NVDA) and Sandisk $(SNDK)$ also in the top ten.

Since its 2017 inception, the Blue Whale fund has seen annualized net-of-fee returns of 19%, beating a respective 12.4% return for its MSCI World Index benchmark, he said. The bottom-up, research-driven fund has nailed early investments and timely exits around the AI trade.

At its start, the fund owned nearly the entire software complex - Salesforce, Adobe, Microsoft, etc, said Yiu. While that trade worked for a while, their complete exit in 2024 saved investors from the selloff to come. He recalled shouting "AI disruption" from the rooftops about software, but said few investors seemed to understand it at the time.

The fund was early to Nvidia (NVDA) buying in 2021 and holding on during a bumpy 2022. The fund's early 2025 investment in SK Hynix was also well timed - the Korean memory maker's shares have surged 685% in a year.

In 2024, the fund picked up Vertiv $(VRT)$ which makes liquid cooling systems for AI data centers, along with Broadcom $(AVGO)$, then Lumentum, the maker of optical interconnects for AI data centers, late last year.

Yiu acknowledges that Nvidia, whose shares are up just over 5% this year, has fallen out of favor. He blames that on unrealistic expectations for Nvidia and Broadcom, whose shares have seen a similar performance this year, to keep delivering blowout rates of returns seen over the past two to three years.

"You cannot expect Nvidia is going to become a $10 trillion company within two years. It might get to $10 trillion, but it would take some time."

But to him, Nvidia's valuation remains attractive, with a forward price-to-earnings ratio of around 22 times, net cash on its balance sheet and heavyweight customers. "When you look at the earnings trajectory it's not stopping at all," yet Nvidia continues to get cheaper.

Apple, by contrast, has a price-to-earnings ratio above 30% while revenue isn't growing.

"I think we are at the very early journey of the AI development or innovation, hence I think there's still going to be a lot of demand for GPUs (graphics processing units)," he said.

Yiu's fund isn't invested in the big hyperscalers, outside of Alphabet, $(GOOGL)$ which they picked up this year on the view it's "trending" in the right direction to be a future winner, with its sophisticated AI model, Gemini.

He said the fund passed on SpaceX $(SPCX)$, whose valuation "has too much of a premium on Elon Musk's magic touch," and will probably be loss-making for the next three years if not longer. He believes OpenAI and Anthropic will also be "too expensive" for them when they eventually come to the market.

Yiu weighs in on the debate around memory, saying he sees that part of AI infrastructure as "the most long lasting."

"The way we understand this is we are going to see a mega explosion of data and content generated by AI," he said. That will need storing, with increased personalized data and ads also ramping up content, he said.

Apple $(AAPL)$ recently raising its prices on some products is proof of just how important memory is to the industry, he said. If an investor believes AI is here to stay, then memory prices are bound to keep rising.

The markets

U.S. stock futures (ES00) (YM00) (NQ00) are mixed, with tech pointing south, as oil (CL.1) (BRN00) drops.

 
Key asset performance                                                Last       5d      1m       YTD     1y 
S&P 500                                                              7483.23    1.70%   -0.93%   9.32%   20.17% 
Nasdaq Composite                                                     26,040.03  2.21%   -3.03%   12.04%  27.69% 
10-year Treasury                                                     4.498      10.30   1.70     32.60   14.70 
Gold                                                                 4078.8     0.92%   -9.41%   -5.85%  22.27% 
Oil                                                                  67.5       -5.55%  -27.35%  17.58%  0.49% 
Data: MarketWatch. Treasury yields change expressed in basis points 

The buzz

Nonfarm payrolls for June are due at 8:30 a.m., with growth of 115,000 expected.

Weekly jobless claims will be released at the same time, followed by factory orders at 10 a.m.

Google (GOOGL) just lost an 8-year battle with the EU over a $4.68 billion antitrust fine.

The bond market is due to close early ahead of the federal holiday on Friday.

SEC probes alleged insider trades that cost Susquehanna millions.

The chart

The chart from JPMorgan shows how hyperscalers have struggled versus semiconductors since last September. Strategist Nikolaos Panigirtzoglou and his colleagues say that unsustainable gap will will start to close if hyperscalers and AI model providers and users see improved monetization and start to capture "a bigger share of the overall AI value-added pie." They warn, though that if analysts' expectations for a sharp deceleration in hyperscalers' capex trajectory from 2027 proves right, "the semiconductor trade could come under severe pressure inducing a more significant and sustained correction in the AI trade in both equity and debt markets."

Top tickers

These were the top-searched tickers on MarketWatch as of 6 a.m.:

 
Ticker  Security name 
TSLA    Tesla 
MU      Micron 
NVDA    Nvidia 
SPCX    SpaceX 
AMD     Advanced Micro Devices 
TSM     Taiwan Semiconductor Manufacturing 
MU      Microsoft 
PLTR    Palantir 
GME     GameStop 
META    Meta 

-Barbara Kollmeyer

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

 

(END) Dow Jones Newswires

July 02, 2026 06:52 ET (10:52 GMT)

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

At the request of the copyright holder, you need to log in to view this content

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Comments

We need your insight to fill this gap
Leave a comment