Every autumn, markets remind investors that patience is not passive — it is a discipline. September has historically been the cruelest month for equities, and this year carries the added weight of geopolitical strain and a weary consumer. Yet within that seasonal turbulence, three companies building the infrastructure of artificial intelligence — Palantir, Marvell, and Snowflake — stand as candidates for those who understand that the price of a thing and the value of a thing are rarely the same.
Three AI Stocks to Buy if September Market Pullback Materializes
Existing customers spending 57% more than a year prior
Mark
Why does September matter so much? Is it just superstition, or is there something real here?
Mimi
It's historical pattern, not superstition. The S&P 500 has averaged a 1.1% decline in September since 1928. In midterm election years, it's been worse—down 5% or more in September-October fifteen times since 1930. This year there's also a war in Iran and consumer spending showing weakness.
Luke
But past performance doesn't guarantee future results. We should be careful about assuming September will pull back just because it has before.
Mimi
Fair point. The article is really saying: if a pullback happens, here are three stocks worth considering. It's not predicting one will happen.
Mark
So these three companies—Palantir, Marvell, Snowflake—they're all expensive right now?
Mimi
Yes. Palantir is at $172.56, Marvell at $221.70, Snowflake at $322.98. The article calls their valuations "pricey" and "frothy." But their growth metrics are extraordinary. Palantir's net dollar retention is 157%—that means existing customers are spending 57% more than they did a year ago.
Luke
How many customers are we talking about? Is 157% retention impressive if the customer base is tiny?
Mimi
The article doesn't specify customer count. It mentions revenue grew 93% year over year, which is substantial, but you're right—we don't know if that's 50 customers or 5,000.
Mark
What about Marvell? What's the actual business there?
Mimi
Optical connectivity for data centers, mostly. As AI workloads have grown, data centers have shifted from copper wiring to fiber optics. Marvell supplies the components. They also license chip design IP to companies like Amazon and Microsoft. Recently they signed a deal with Alphabet for products that work with Alphabet's TPU ecosystem.
Luke
How much is that Alphabet deal worth?
Mimi
An analyst at Piper Sandler estimates it could reach $18 billion in annual revenue once fully ramped. But that's one analyst's projection, not a guarantee. And it doesn't start contributing until fiscal 2029.
Mark
So we're betting on future revenue that hasn't materialized yet.
Luke
Exactly. The article projects 50% revenue growth in fiscal 2028 and points to the Alphabet deal as a driver. But projections are not facts. We should know that distinction.
Mimi
True. But the company does have real customers today—Amazon, Microsoft—and a real product in optical components that the industry needs.
Mark
What about Snowflake?
Mimi
Cloud data warehousing. They separate storage from compute, so customers can store data once and process it across multiple cloud providers. Last quarter, revenue was up 35%, net dollar retention was 126%. They've launched two new AI products—CoCo, a coding agent, and CoWork, an agentic app.
Luke
Are those products generating meaningful revenue yet, or are they still early?
Mimi
The article doesn't say. It mentions they're seeing "rapid adoption," but doesn't give numbers.
Mark
So again, we're looking at potential, not proven results.
Mimi
Right. But the potential is real. These are companies with demonstrated ability to grow revenue and expand customer spending. The question is whether their current stock prices reflect that potential or overshoot it.
Luke
And that's why the article frames them as buys on a pullback, not buys at current prices.
Mimi
Exactly. If prices fall, the risk-reward shifts in your favor.
Der Puls
- September's historical average decline of 1.1% in the S&P 500 — steeper in midterm years — is arriving this cycle alongside an Iran conflict and visible consumer stress, sharpening the risk of a meaningful pullback.
- Palantir, Marvell, and Snowflake each carry premium valuations that make them vulnerable to broad market selling, yet their underlying business momentum — triple-digit revenue growth, net dollar retention well above 100% — makes a sharp decline feel more like dislocation than deterioration.
- Palantir's ontology-based AI platform is reducing the hallucination problem that undermines less grounded systems, and its 157% net dollar retention signals customers are not just staying — they are accelerating their commitment.
- Marvell sits at the physical heart of AI infrastructure, supplying the optical components that replace copper in data centers, with an Alphabet deal that analysts project could eventually yield $18 billion in annual revenue.
- Snowflake's model-agnostic, storage-compute-separated architecture has made it essential plumbing for agentic AI deployments, with management guiding revenue growth to accelerate toward 38% next quarter.
- The thesis converges on a single idea: if September delivers its historical correction, long-term investors may find that fear has briefly priced these companies as though the AI decade were already over.
Every autumn, markets remind investors that patience is not passive — it is a discipline. September has historically been the cruelest month for equities, and this year carries the added weight of geopolitical strain and a weary consumer. Yet within that seasonal turbulence, three companies building the infrastructure of artificial intelligence — Palantir, Marvell, and Snowflake — stand as candidates for those who understand that the price of a thing and the value of a thing are rarely the same.
September has a reputation earned over nearly a century of data. Since 1928, the S&P 500 has fallen an average of 1.1% during the month, with the damage deepening in midterm election years — the index has dropped 5% or more during the September-October stretch fifteen times since 1930. This year, an ongoing conflict in Iran and a strained consumer sector add further weight to the seasonal pattern. For investors with conviction and a longer horizon, however, pullbacks are not simply losses to endure — they are entry points to consider.
Three artificial intelligence companies stand out as potential buys if markets do stumble. Palantir Technologies, trading near $172 with a $416 billion market cap, has built something closer to an operating system for AI than a conventional software product. Its platform organizes data into a structured ontology — linking information to real-world objects and processes — which reduces the costly hallucinations that undermine less sophisticated systems. Revenue climbed 93% year over year last quarter, and net dollar retention reached 157%, meaning existing customers are dramatically expanding their usage. A retention rate above 100% is uncommon; 157% is exceptional.
Marvell Technology, near $221 per share, has positioned itself at the physical layer of AI infrastructure. As data centers have shifted from copper to fiber optic connectivity to handle AI workloads, Marvell supplies the digital signal processors, switching equipment, and broadband components that make that transition possible. The company counts Amazon, Microsoft, and Alphabet among its major customers, and a recently signed multi-year agreement with Alphabet — covering inference accelerators and storage controllers within Alphabet's tensor processing unit ecosystem — could generate an estimated $18 billion in annual revenue once fully ramped, according to one analyst. The company projects 50% revenue growth by fiscal 2028.
Snowflake, near $323 per share, dominates cloud-based data warehousing through an architecture that separates storage from compute, allowing customers to process data across multiple cloud providers without duplication. That flexibility has made it foundational infrastructure for companies deploying agentic AI systems. Revenue rose 35% last quarter, net dollar retention reached 126%, and two new products — CoCo, an AI coding agent, and CoWork, a ready-to-use agentic application — are gaining early traction. Management expects growth to accelerate to 37–38% next quarter. By remaining model-agnostic, Snowflake has avoided the trap of vendor dependency and built a durable position in an emerging market.
None of these companies is inexpensive by conventional measures, and that is precisely the point. If September delivers the correction that history suggests is plausible, these three stocks may briefly trade at prices that reflect the market's anxiety rather than the underlying trajectory of artificial intelligence — and for patient investors, that distinction is everything.
September has a reputation. Since 1928, the S&P 500 has fallen an average of 1.1% during the month, and the pattern worsens in midterm election years—the index has dropped 5% or more during the September-October stretch fifteen times since 1930. This year, the headwinds feel particular: a war in Iran and a consumer sector showing strain add weight to the seasonal weakness. But market pullbacks, for investors with conviction and a longer view, can be moments of opportunity rather than panic.
Three artificial intelligence stocks stand out as potential buys if the market does indeed stumble in the coming weeks. Each has built real business momentum, yet each trades at valuations that would look far more reasonable if prices fell sharply.
Palantir Technologies trades at $172.56 with a market capitalization of $416 billion, and its valuation is the first thing skeptics point to. But the company has positioned itself as something closer to an operating system for AI itself. Its platform ingests data from multiple sources, organizes it into what the company calls an ontology, and links it to physical objects, processes, and concepts—essentially grounding artificial intelligence in the real world in a way that reduces the costly hallucinations that plague less sophisticated systems. The company's AIP product line has resonated with customers across industries. Last quarter, revenue climbed 93% year over year, while net dollar retention hit 157%—a figure that signals existing customers are spending significantly more than they were a year prior. Any retention rate above 100% is rare; 157% suggests the company is not just keeping customers but expanding their usage dramatically.
Marvell Technology, trading at $221.70 with a $192 billion market cap, has positioned itself at the center of a fundamental shift in data center architecture. As AI workloads have grown, the industry has moved away from copper wiring toward fiber optics for connectivity between servers. Marvell supplies the optical components that enable this transition—digital signal processors, switching equipment, and broadband analog components. Beyond that, the company licenses intellectual property to help other firms design custom chips. Amazon and Microsoft are major customers; more recently, Marvell signed a multi-year agreement with Alphabet to supply products that work within Alphabet's tensor processing unit ecosystem, including inference accelerators and storage controllers. An analyst at Piper Sandler estimates this Alphabet deal could eventually generate $18 billion in annual revenue once fully ramped. The company projects 50% revenue growth in fiscal 2028, with the Alphabet contract beginning to contribute in fiscal 2029.
Snowflake, at $322.98 per share and a $117 billion valuation, dominates cloud-based data warehousing and analytics. Its architecture separates storage from compute, allowing customers to store data once and process it across multiple cloud providers—a flexibility that has made it essential infrastructure for companies deploying agentic AI systems. Last quarter, revenue rose 35%, driven largely by AI solutions, while net dollar retention reached 126%. The company has launched two new products gaining traction: CoCo, an AI coding agent, and CoWork, a ready-to-use agentic application. Management expects revenue growth to accelerate to between 37% and 38% next quarter. By positioning itself as model-agnostic—compatible with any AI system rather than tied to one vendor—Snowflake has created a durable moat in an emerging market.
Each of these companies has demonstrated the ability to grow revenue at rates most mature businesses can only dream of, while also expanding their customer base's spending. The risk, of course, is that none of them is cheap by traditional measures. But if September delivers the pullback that history suggests is possible, these three stocks would offer entry points for investors who believe artificial intelligence will reshape business over the next decade.
Bemerkenswerte Zitate
Any number over 100% indicates growth from existing clients of over a year, and that type of number is rarely seen.— Source analysis of Palantir's net dollar retention
Piper Sandler analyst David O'Connor believes this could be an $18 billion annual revenue deal once the program is fully ramped up.— Regarding Marvell's Alphabet contract
Möchten Sie die ganze Geschichte?
Das Original lesen bei The Motley Fool ↗
Economy & Finance