Rhys Williams, chief investment officer at Wayve Capital Management, told Phil Rosen on Full Signal that today's AI boom is not a replay of the 1999 dot-com bubble.
Williams, who has been in the market since 1987, laid out a barbell strategy for 2027 that pairs AI infrastructure stocks with health care and energy.
Why this isn't 1999
"In 1999, Cisco was trading at about 120 times earnings," Williams said. The 10-year Treasury yield ended that year above 6%. Today, he said, Nvidia trades at about 15 to 16 times earnings and Micron at about six times.
"Sentiment is awful," he said, even though "fundamentals are very strong." He blamed political backlash against data centers, and said power and water concerns have largely been addressed.
Sentiment selloffs and the cracks to watch
Williams compared today to the internet era. From 1994 to 2000, he said, the semiconductor index rose 850% despite two 50% corrections, and the companies never missed earnings.
In December 1999, a Taiwan supplier told him Broadcom had cut orders 40%, months before the Nasdaq peaked. He doesn't see those cracks now. "I think this is just a buying opportunity," he said.
He also said he was impressed by how little cash Anthropic has used, based on its S-1. The filing showed $131.6 billion in cash and short-term investments as of June 30, after the company raised $141.1 billion.
The barbell portfolio for 2027
"I do think a barbell approach makes some sense," Williams said, pairing AI infrastructure names with sectors that tend to move opposite AI.
In health care, he named Ligand Pharmaceuticals and Eton Pharmaceuticals as smaller companies with "a lot of shots on goal." He'd inch into Ligand after its run rather than take a full position.
He called Takeda, the large Japanese drugmaker, a value pick that pays a nice dividend. "This is, I think, a good sleep at night stock," he said.
Energy as the AI hedge
On the days energy goes up, AI goes down, Williams said. If the wars involving Russia and Iran drag on, he expects oil prices to rise, and he sees Exxon and Chevron as the conservative way to play it.
He also likes refiner Phillips 66 because diesel and jet fuel prices are high, and the Alerian MLP ETF, whose dividend he put at around 7%.
His AI stock picks
Williams said Astera Labs gets about 75% of its business from Amazon, which he said makes it nearly a pure play on AWS. He said Amazon is the less risky way to play that trend, while Astera may offer more upside.
One risk he flagged is whether Astera keeps its content in Amazon's Trainium 4 chip, which he said arrives at the end of 2027.
His base case is no recession, and he thinks health care will be a big winner, in line with Adam Parker's bullish call on health care.
Tune in to the full conversation with Rhys Williams on YouTube, Spotify and Apple Podcasts.
For more on the AI cycle, read J.P. Morgan's Stephanie Aliaga on Why the AI Cycle Is Still Early and the Sectors She Likes, and subscribe to the free Opening Bell Daily newsletter for a daily briefing on stocks, macro and Wall Street.
This article is for informational purposes only and is not investment advice. Views and stock opinions are those of Rhys Williams and market figures are as stated by the guest at the time of recording.
