The State of the AI Trade

By David Moenning

Don't look now fans, but the AI trade - you know, that high-octane corner of the market loaded with new-fangled semiconductor chips, cloud plays large and small, and those mind-bending large-language models - is still kicking up plenty of dust in the market. The question, of course, is whether the trade, which has been in vogue for going on 3 years now, is something we should continue with.

There can be little argument that we've definitely seen some bumps lately, what with volatility spiking and many of the big names taking some pretty healthy hits – even after posting sometimes spectacular earnings.

The bears are quick to argue that the trade is over, with some even shouting fraud at every turn. Our furry friends argue that valuations are reminiscent of the dot-com era and that investors need to get out now.

But in my book, this sector's got legs for the long haul, even if the short-term ride requires investors to remain in their seats with seatbelts securely fastened.

To be sure, it's a tale of two camps: the bulls charging ahead with visions of world-changing tech, and the bears growling about bubbles and overreach. So, this morning I thought it might be a good idea to break down the arguments from both teams.

The Bull Case: Why AI Can Keep Movin' On Up from Big Picture Perspective

The primary point from the bull camp is that adoption in AI is just getting started and appears to be snowballing. Everywhere you turn, there it is: Click here to try our AI model.

In addition, our heroes in horns remind us that we must understand that "AI" is SO much than chatbots and LLMs (large language models such as ChatGPT, Gemini, Claude, Grok, etc.).

AI is NOT a product companies buy off the shelf. As such, it will take time for companies to figure out the best way to utilize the technology. The bulls see a supercycle in the making for computing. And frankly, the evidence is stacking up. Here's my quick rundown on what’s got the optimists still excited about the future:

• Adoption is Exploding: API calls and token usage for the heavy hitters - think OpenAI, Google, Meta - continue to skyrocket. It's not about reinventing the wheel with fancier models; it's the everyday integrations into business workflows that can juice revenues and cement moats. This stuff's accelerating, and from my seat, that spells opportunity.
• Productivity Lift for the Win: These large language models aren't just hype; they're cranking up what’s called total-factor productivity across the board. That softens economic headwinds and keeps liquidity flowing, turning AI into a macro powerhouse that spills over into cloud, compute, and beyond. If you're playing the broader perspective, this is your tailwind.
• Not a Bust: These recent selloffs? Yeah, post-Oracle (ORCL), Palantir (PLTR) and Broadcom (AVGO), it's been dicey. But my take is this was more of a rotation or stress test than a meltdown - capital appears to be shifting to the infrastructure backbone. The ecosystem's maturing, not crumbling, and that's bullish for the long-term, especially in the big "picks and shovels" players.
• Policy Tailwinds: With calls for federal backstops on AI capex (looking at you, OpenAI), and a dash of economic nationalism thrown in, this could squash financial squeezes and build out national champs. Dips? I’m of the mind that they're buying ops in this environment.
• It's a "Supercycle:" AI's disruption is creating waves of innovation, and we see the sweet spot is in the plumbing - compute, data, storage. Partnerships and open-source momentum keep the narrative alive, with short interest staying low on the mega-caps. Institutions are all in, and that outweighs the wobbles.
• Metrics Don't Lie: Solid earnings and cash flow backs the leaders, and with broader tech exposure buffering pure AI bets, the risk-reward suggests upside, well for me, anyway.

In case you've been living under a rock, this isn't pie-in-the-sky stuff; it's grounded in real trends. Real earnings and real revenues. Sure, Ms. Market often throws curveballs, but the bulls argue - and I tend to agree - that going against this movement could be a mistake for portfolios.

But... Bear Complaints are Worth Noting

To be fair, the skeptics aren't just whistling Dixie here. We can all agree that valuations are stretched, capex is through the roof, and competition's heating up. Those seeing the glass half empty (or bone dry) point to some real risks that could derail the party.

Here's the bearish narrative in a nutshell:

• Bubble Territory: It's a sentiment tug-of-war right now, with the likes of Oracle's debt worries and overvaluation fears at the forefront. Many big names have publicly drawn bubble parallels, warning that if hyperscalers pull back on spending, meaningful corrections could be imminent.
• Moats Can Crumble: The U.S. LLMs may not hold their edge forever - Chinese outfits are cranking out cheap, open-source winners that leapfrog leaderboards and slip into tools like Perplexity. Commoditization could turn this into a race to the bottom, with marginal costs killing profits. Ouch.
• Capex Hangover: All that cash flooding into GPUs, TPUs, RAM, and data centers risks building ghost towns if demand (and/or energy availability) doesn't keep pace. Add in talk of GPU depreciation jitters and export curbs to China, and you've got short-term pain that could hobble the whole space.
• Credit Crunch? The bears argue that a credit crunch is looming as the big players start to issue debt to pay for the buildouts. The argument goes like this: if credit tightens, capital costs will spike, and deals could easily fall apart - cue the degrossing.
• Near-Term Rotations Sting: Long-term transformative? Absolutely. But the hype wave appears to be receding, with rotations out of semis and unmonetized plays (hello, OpenAI ads) weighing on "the trade." No full-blown panic yet, but selective pressure has definitely been building - especially in the more speculative areas.

The bottom line is only time will tell which team will win the day. But unless fundamentals flip, I'm not jumping ship just yet. My plan is to keep an eye on the data and try not to fight the primary trend.

View Blog Post