Why I'm intrigued by Anthropic's IPO

As investors, we should always be searching for the point of integration, or the chokepoint, in a value chain. It’s the place where the most value accrues.

On the internet, that’s $Alphabet(GOOGL)$ .

In tech hardware, it’s $Apple(AAPL)$ ’s iPhone.

On PCs, it was $Microsoft(MSFT)$

This is the point where lots of suppliers are forced to build for the integrated point, modularizing their businesses. And customers make their purchasing decisions based on this integration.

I think $Anthropic(ANT)$ has a chance to be the point of integration in AI, and that’s what makes it an interesting business today. I’m not ready to make it a buy, but it’s on my radar and this is what could make it a 10x stock.

The AI Operating System

When chatbots were first launched, I was skeptical that the chatbot itself was a point of integration, as some thought. A chat in ChatGPT isn’t all that different than a chat in Gemini or Claude or any other chatbot, and there was no real integration with other things you may be doing, despite OpenAI trying to make plug-ins a thing.

But over time we’ve seen the industry begin to take shape around what could be a new future, and chatbots are starting to modularize their supply.

Here’s how the thesis goes:

Claude becomes the application users interact with across all hardware. It’s where you may go to search for an answer or find a recipe at home. It’s also where you go at work to make an image or video, or maybe interact with a work software application. It may even be the future of office productivity products.

The sign that this could be the future is how quickly companies have launched MCPs, or Model Context Protocols. In essence, an MCP allows you to use an application or service you have through Claude, OpenAI, or another chatbot. Salesforce has an MCP, so does ServiceNow, and even Workday.

You can see how users (and companies) could find this appealing. No more “where do I put my vacation time?” or “how do I get reimbursed for lunch again?”

Just access the tool in the MCP and ask Claude to do it for you.

For example, I use Fiscal AI’s MCP to pull data and create my own graphics. I’ve historically used their charts like this one 👇

Now, I can simply use Claude or ChatGPT (which I’m using here) to make a custom image with Fiscal AI data and even SEC filings with more granular data.

Using this prompt:

I get this:

This is a small example, but now extend this to dozens of pieces of software and systems you may use at your job every day. Would it make sense to just plug into a dozen MCPs and do it all in Claude?

Windows didn’t win because it had the best operating system. It won because it had the most-used operating system, and that created a network effect that attracted more users and more suppliers. The draw here may be that Claude has the best models and they’re plugged into the chat interface users prefer to use. Suppliers are already coming to the platform, so if these trends continue, Anthropic grows.

Anthropic’s Challenge

The theory I’ve laid out is the bull case for Anthropic, and if it plays out that way, this is a 10x company, no doubt.

But Anthropic doesn’t operate in a vacuum.

ChatGPT exists.

Google exists.

Apple exists.

All of these companies are going to try to be their own point of integration. And they have a good shot at it.

I have the most questions about the future of AI hardware and what that looks like. Is Apple iOS or Google’s Android able to simply plug their AI into new hardware, and it’s game-over for Anthropic’s integration dreams?

Or will hardware makers be modularized and the only application that matters is the AI you choose to plug into?

The case for MCPs I made above also extends to ChatGPT and Gemini. If users choose them, Anthropic is just another chatbot.

There are also the financial questions. Currently, Anthropic and all of the AI labs are on a model training treadmill that seems like it’ll never end. Unless model development spend as a % of revenue falls, there will never be operating leverage in the business.

Why Anthropic Could Be Different

As I began with, I’m not ready to buy shares of Anthropic yet, but I’m starting to formulate the bull case for the stock in case the price gets more attractive.

After the IPO, Anthropic has about $200 billion on the balance sheet to weather any storm and build out the product line horizontally that could make Claude more of a platform than it is today. Office tools are a move in that direction, but more is coming.

It’s not guaranteed that AI is a “winner-take-all” market the way many tech markets are, but if it is, this is the leader in the clubhouse right now.

With that much upside, it’s OK to be late to the market’s biggest winners. Microsoft, Apple, Amazon, and Meta were the ultimate winners in their fields, and even buying their IPOs years after they went public was a profitable buy.

The time opportunity may be compressed in a world of AI, but it doesn’t mean I want to rush into an IPO, and Anthropic’s drop post-IPO shows why. But the stock is on my radar.

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Report

Comment

  • Top
  • Latest
empty
No comments yet