# Why I, a dedicated small-cap investor, am finally buying the Alphabet (GOOGL) dip after Q3 earnings — retained record

> Retained reddit record from reddit:ValueInvesting.

Canonical: https://highsignal.app/data/records/cb41cb4b3509d7ad

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:ValueInvesting)  published 2026-10-30 07:34 ingested 2026-10-31 02:40

[original source ↗](https://www.reddit.com/r/ValueInvesting/comments/1xj8ysr/why_i_a_dedicated_smallcap_investor_am_finally/)

## retained text



```
Alphabet just dropped its Q3 2026 earnings, and the market completely panicked. The stock tanked nearly 10% after hours and closed the next regular session down a massive 12.6%. For a mega-cap sitting above a $3 trillion valuation, shedding over 10% in a single day is a rare sight.

The sell-off basically boils down to two things. First, Alphabet spent so aggressively on data centers and AI infrastructure that its free cash flow (FCF) flipped to negative $9.1B. Second, Search revenue growth decelerated from 17% in Q2 to 12% in Q3.

However, after digging into the actual numbers, my takeaway is the exact opposite of Wall Street's knee-jerk reaction. Short-term price action aside, this was a phenomenal quarter for long-term holders. In fact, as someone who almost exclusively invests in small- and mid-cap stocks, this report made me seriously want to open a position in GOOGL. Here is why.

**Why I usually avoid Big Tech, and how Google solved the reinvestment problem**

I usually hunt for small- and mid-cap compounders with wide moats because I want to beat the market. Basic math dictates that the bigger a company gets, the harder it is to maintain high growth rates. To keep compounding, a business needs to reinvest its cash flows back into high-ROIC opportunities. Once a company reaches mega-cap status, that reinvestment runway usually dries up, leaving buybacks and dividends as the only real levers left. That is why I generally stay away from Big Tech.

Looking at Google right now, this massive cash cow has found a massive, high-ROIC reinvestment runway. For the past few years, Google has been pouring an almost unfathomable amount of money into CapEx—$51.4B in Q3 alone, and roughly $200B guided for the full year—all aimed at data centers and AI infrastructure.

When Berkshire Hathaway started building its massive stake in Alphabet, Warren Buffett compared the current AI CapEx cycle to building the railroads. Laying tracks requires astronomical upfront capital, but once the infrastructure is in place, latecomers simply cannot compete. Google's data center buildout is the exact same game. It is a hyper-scale, capital-intensive moat where you have to spend tens of billions just to sit at the table. Can smaller companies or late entrants catch up? Not a chance.

**Cloud is printing money, and Search is far from dead**

The best part is that Google isn't burning cash on a pipe dream—the returns are already showing up in the numbers. Google Cloud was the undeniable star of this report. Cloud revenue surged 103% YoY to $30.7B (literally doubling in a year), and operating margins hit a whopping 36.2%. Meanwhile, their Cloud backlog sits at an insane $561B. Every dollar they pour into CapEx is translating directly into high-margin revenue in an arena where smaller players can't even afford the entry ticket.

Search was honestly my biggest concern when generative AI first blew up. Ever since ChatGPT and Claude came out, my own search habits changed. I completely stopped using local search engines, and I found myself using Google Search less frequently too.

Yet, Google adapted faster than anyone by integrating AI directly into Search. Despite the market freaking out over "deceleration," Search revenue still grew 12% YoY at a massive scale, and total search query growth actually hit a 5-year high. Everyone has been writing the obituary for Google Search for two years, yet it continues to print money and grow.

**Not #1 in every category, but the only fully integrated empire**

Of course, Google isn't the only hyperscaler in town. AWS and Azure still hold larger market shares in cloud, and if we are being completely honest, Gemini isn't decisively beating ChatGPT or Claude on pure model performance. And when it comes to AI chips, Nvidia is still the undisputed king. Looked at in isolation, Google isn't #1 in every single vertical.

However, Google is the ONLY company on the planet that owns the entire AI value chain from top to bottom. They have a frontier model (Gemini), custom silicon (TPUs), global cloud infrastructure, and the world's largest proprietary training data and distribution ecosystem (Search, YouTube, Android). They are vertically integrated and rank near the top in every single layer of the stack.

This vertical integration is going to create an insurmountable cost advantage. If you look at the pace of LLM development, whenever a frontier lab releases a state-of-the-art model, competitors catch up to that previous tier within months. No single lab has a permanent, insurmountable moat on model intelligence alone. If top-tier models become commoditized and perform similarly, the ultimate winner will be decided by unit economics. When performance is comparable, enterprises and consumers will simply choose the cheapest option—because everyone wants to run as much AI compute as possible.

**TPUs: The ultimate weapon against the "Nvidia tax"**

This is where Google's custom TPUs change the math. While other hyperscalers are forced to pay massive margins to Nvidia for GPUs, Google can drastically lower its compute and inference costs using in-house silicon, effectively eliminating its reliance on Nvidia.

What stood out most on the earnings call was the external demand for TPUs. Roughly 100 major enterprise customers—including Anthropic (makers of Claude) and Apple—are already running on Google's TPUs, and Google even started generating direct revenue by selling TPU systems to players like Oracle. Demand for Google Cloud and TPUs is so absurdly high that Google doesn't even have enough capacity to meet external orders while training its own models. They are literally leaving money on the table purely due to supply constraints.

**TL;DR: Good negative FCF vs. Bad negative FCF**

Seeing a cash machine like Alphabet post negative $9.1B in quarterly FCF is jarring at first glance. If they were burning that cash on speculative moonshots with no visibility, I would run for the hills. But they are pouring that capital into laying the tracks for a business currently printing 36% operating margins with a $561B backlog. As a long-term investor, I would much rather see Google aggressively reinvest to build an unassailable infrastructure moat than sit on a pile of cash doing buybacks while growth stagnates.

With the market wiping out 12.6% of Alphabet's market cap over short-term FCF fears, this dip looks like a rare opportunity to climb on the back of a giant.
```



document fetched 2026-10-31 00:48 · canonical https://reddit.com/r/ValueInvesting/comments/1xj8ysr/why_i_a_dedicated_smallcap_investor_am_finally
