AI—Parnassus Investments’ Todd Ahlsten asks where are we now?

Sharing his latest analysis into AI from an investor’s perspective, Todd Ahlsten, Chief Investment Officer, Parnassus Investments and Portfolio Manager, ABN AMRO Parnassus US ESG Equities, tells us why he believes that emphasis should remain on those quality companies with strong cash flows and reasonable valuations.

A year ago I suggested there were two economies: the AI economy and the rest of the world. The meetings I had with tech company leaders in Silicon Valley last spring were buzzing with talk about the amount of investment being made in AI based on the free cashflows the tech companies were generating. At that time, the economic uncertainty resulting from higher interest rates seemed like a distant risk. Ultimately, it was, as we made it through last year with a fairly strong economy.

This year, tech giants running hyperscale datacentres have committed to investing $325 billion in AI, even as questions arise about what the return on that spending will be. We’ve seen the momentum behind AI stocks unwind a little this year amid the broader market volatility, but I’m not seeing any yellow flags yet on AI spending. Most of the investment in AI datacentres has been for training the AI models, but I believe the compute needed to support AI inferencing and reasoning could cause an acceleration in investments in the next few years.

I’m keeping an eye on other AI trends, such as the shift toward custom silicon. Nvidia’s success in manufacturing general-purpose semiconductor chips (GPUs) for video, 3D and AI processing may face competition from custom silicon chips tailored for specific applications. With custom silicon, companies can manufacture chips for the tasks and processes they need, with greater cost and energy efficiency. This shift could allow companies like Broadcom to share a portion of the market Nvidia now dominates.

Finally, there is the ongoing debate around how much AI providers will be able to monetize AI. Enterprise-level adoption has been a bit slower than expected. In the past, transformative tech innovations like the internet and cloud computing were not overnight changes. They took years or even decades to fully monetize. AI is likely no different.

These risks are worth watching, but as I said, my long-term view on AI remains positive, and I continue to believe in its transformative potential. We’re overweight the hyperscalers that can monetize workloads and improve the transition from on-premise systems to cloud computing. We are also overweight the shift to custom silicon and the long-term monetization of AI, especially by leading software companies. Looking at a wide range of outcomes with AI, companies like Microsoft, Amazon and Alphabet will remain quality companies that are central to the economy. They are some of the best companies in the world, with strong cash flows and reasonable valuations.

There’s an old saying that capital flows to where it’s treated best. Investors seek the highest returns and the best payback, whether in public or private markets, fixed income or equities. Over the past century, history has shown that the U.S. stock market has outperformed the vast majority of asset classes. And in the long run, I believe quality companies will be the ones that help deliver those returns.