Meta 为持有评级；受益于 AI 相关举措，营收将持续保持十几个百分点区间的中段增长。

So we're gonna be talking about one company that is now one of the largest positions in the top five largest positions in our portfolio here at Chip Stock Investor, Meta.

Meta has definitely seen its share of turmoil. There's been a number of lawsuits. They've settled some of the lawsuits … they've settled some of them. They've no doubt have many more to come.

They're just universally hated, and yet they have 3.6 billion daily active users across their family of apps every day. Facebook, Instagram, and WhatsApp primarily.

And now, Meta Muse is creating some buzz. Let's talk about Meta Muse and why this company stays a hold in the CSI portfolio.

the way we used fiscal.ai to break this down is the family of apps revenue in purple, total of $60 billion per quarter, and then a sub-segment of that is the advertising revenue, which comes in at about $59 billion.

What's not pictured here, which we're gonna show you a chart in just a moment, is the Reality Labs revenue, which is about $400 million a quarter, minimal amount of revenue.

So the vast majority of the revenue each quarter comes from advertising for Meta. Reality Labs is excluded. That means there's less than a billion in revenue. Let's say approaching a billion dollars in revenue generated by the family of apps part of the business that is not advertising.

This comes from business services as well as subscriptions, which is a newer revenue line item, but obviously, that is close to irrelevant.

Meta Muse is the latest attempt to try to change that.

Now, Kasey, you did mention Reality Labs, over 400 million in quarterly revenue in Q2, 430 million, roughly. It is still growing, but look at the cash that they burned through to this day.

The Family of Apps, social media, ridiculously profitable. Their vertical integration of their ad software built atop Meta's own infrastructure is most definitely the real deal.

Reality Labs is a drag, still burning through billions of dollars every quarter. It sounds like they are going to kind of rein this in, which they've talked about numerous times over the years.

At this point, Metaverse is still a dirty word. It's still very much a dirty word to investors. But I think a little context is helpful here because there's a reason why Meta has made multiple attempts at some sort of hardware business, and it all comes down to the fact that they are a software company, and especially a mobile software company, social media primarily consumed on mobile devices, on smartphones.

And as a result, they are highly reliant on Android, Alphabet's Android mobile operating system, and Apple iOS. If you've been invested in Meta for a long time, you probably remember the iOS 14.5 app tracking transparency data opt-out that was a pretty big impact for Meta in 2022.

Bad timing with the last bear market, with the Fed rate hike cycle. And so for all the hate that the metaverse gets, you know, AR, VR, helmets, smart glasses, seems like very much a love it or hate it thing with a lot of average consumers probably falling in the no thanks camp.

The reason for that is because they're reliant on someone else's hardware ecosystem, they're constantly having to try to circumvent disruption to their primary source of revenue.

And that's why they've made such a push into hardware, like data centers.

That's one of the things that's made Meta very durable through all of this. So they've gone in both directions, and obviously the data center part of that has been wildly successful.

Most of the time unbeknownst to investors, Meta's very much a data center business. Has been for the life of the company being publicly traded.

Now that's very well known, but they can do more. And as shareholders, we would expect them to do more, to continue to develop and evolve their business, and AI most certainly presents an opportunity for them to do just that.

We discussed this last year. We had a series of articles over on Semi Insider about the web browser wars. The web browser has made a comeback in the AI era. It's very much a medium that is in high demand, a way that software is consumed in the AI era, be it an LLM or if you're interacting with AI agents.

And we actually listed at that time Meta in the top spot as a company that would benefit from that because they also did this, a while back. They have apps like Facebook and Instagram running on a dedicated web browser, and Meta Muse, the AI agent, is also doing that as well.

It's another way that they can sort of circumvent the hardware distribution layer without having to directly distribute their own hardware and try to get their own hardware in the hands of consumers.

A web browser kind of achieves some of that, if you can keep the consumer within your ecosystem that way.

Let's show you another chart from fiscal.ai, the capital expenditures versus the total revenue that Meta has been receiving. As you can see, this chart goes back to 2015. They have been ramping up CapEx spending, there in pink, $89 billion in the last 12 months alone.

Huge infrastructure build-out for these data centers and vertically integrating their software and hardware.

And this is not new news from us that CapEx has completely eaten up free cash flow. We'll get to that in just a moment. But the big question now, or the big question has been, besides supporting Facebook's social media apps and WhatsApp, what else is that CapEx going to go to?

Meta Muse is one of those things. It is built atop the same data center infrastructure as the apps.

And the idea with this AI agent is to create a type of aggregator. Think of it almost like a personalized aggregator, kind of like if you were checking the price of flights or hotels or using Google to compare the price of the same product across different e-commerce sites.

In this case, it's automated. An AI agent, I think we can kind of strip away some of the jargon and just call it that. It's your own personal aggregator tool. It can connect to your email and other tools to help it understand context for you, what's most important to you, go get jobs done that you ask it to go get done, present you with reports on that.

The list is virtually endless here, but that's essentially what it is. Built atop the same infrastructure.

This is the way Meta could potentially start to slowly, gradually diversify its revenue beyond just advertising, because Meta Muse is going to be sold via subscription beyond a free tier. It's not gonna be free for everyone forever.

There's a few ways that we believe Meta is going to be able to realize revenue from this Meta Muse project. Consumer subscriptions, which Nick just mentioned, those plans start for some of these users at 7.99 for the core plan per month, and then premium for 19.99 a month.

And those business packages coming in at a much higher rate for that product. There's also business agents, APIs, and enterprise tools, commerce and performance fees.

This is going to reemerge again as it did in the past, vertically integrated walled gardens versus more horizontal aggregation tools. We'll see how this pans out.

And of course, Meta is going to be capitalizing on their huge ad market. Even the free tier of Meta Muse is definitely going to have some ad tracking, so if you're using Meta Muse, you're helping bring in some revenue for Meta.

And the value of an ad like that would be theoretically much higher because they would be tracking not just your activity, but they'd have a much more solid read on your intent.

If you told Meta Muse, "Hey, go search for plane tickets," it could probably figure out via your email, inbox and the directions you just gave Muse, this user might be going on a business trip.

There would be a lot of advertisers interested in that information and that would pay a higher price to place an ad in front of you if they knew that information.

To figure out the best case and worst case scenarios or maybe just the base case. So let's consider all of the cases here. Could this blow up the business or could it be a huge success?

On the bear case side, let's start with direct revenue. That would be from subscriptions. I think it's a little bit easier to track the uplift in ad revenue, average revenue per user.

That's going to be more of the so-called black box. We're not gonna have a lot of ability to measure what's happening there. But on the direct revenue, that's pretty easy. The tier of the subscription multiplied by the number of users.

The bear case on the left side, that direct revenue subtotal of three and a half billion annual sales is less than 1% of Meta's 3.6 billion users signing up for a Muse account, and minimal amount of businesses using Meta Muse for their marketing efforts, their business management, the millions of businesses that use WhatsApp every day. Minimal uptake there.

The base case is something like 2%-ish. And then on the right, the bull case, that would be three, four, or higher percent of the 3.6 billion users adopting Meta Muse and a lot more business consumption especially.

That would be the biggest item there, business agents, APIs, so using some of the Meta AI models within those enterprise tools, and then a fair amount of performance fees and e-commerce fees.

Being logical and thinking investors here at CSI, I think we can safely say we'd put this somewhere between the bear and the base case. I think less than $20 billion per year in revenue after a year or two would be a safer assumption, and let's be surprised to the upside if not. Let's sandbag.

Total revenues, we've already showed you what Meta is raking in on a quarterly basis, and then the percent change is the little line there, the little pink line that you see. Over the last few years, they've definitely ramped up that revenue year over year.

We believe that this revenue growth can continue maybe in the mid-teens percentage with these new changes that Meta has implemented.

But increasing revenue has not been the problem for Meta. They have been able to grow revenue sustainably, but one of the biggest things that we see as a risk is trust.

Yeah, and that's where the e-commerce and performance fees are going to come in, and also will consumers and businesses give more of their time and attention and development efforts to Meta's platform?

The CapEx story makes the valuation tricky. The revenue growth is there so far, despite all the controversy. We can say that on average, consumers must trust Meta enough. They continue to grow their user count.

Businesses are still using it. But that could change in the AI agent era, and then the big question on top of that is all of the CapEx they've been spending, does that end up being a total waste if a bunch of walls start getting thrown up around other competing services that Meta Muse would like to access to make it useful in the first place?

So not just trust from Meta's average user, but also trust from all the other services that a consumer would like to use, and that's where some serious roadblocks are going to get thrown up.

Can Meta continue to have EBIT margins and operating margins in the 30% or better range?

There's going to be at least some period of time, it appears it has started now, where that CapEx depreciation on the data center assets, on the servers from Nvidia, CapEx on Meta's own custom chip design like MTIA, working with Broadcom and others on those, working with TSMC to manufacture them, there is going to be a dip.

But does that ramp back up as some of these new projects start to get monetized?

We won't dig into all of the different profit margin metrics specifically. We're just gonna keep it simple here, and we're gonna do earnings per share on a GAAP basis. What would have to happen over the next five years to make Meta a good, fair price right now at, as of this recording, after Monday, September 21st, market close, over $740 per share.

The market definitely thinks Meta Muse is a win automatically, so that increases the five-year CAGR for earnings per share up to 16%. Not as reasonable as it was on Friday, certainly not as reasonable as it was six months ago and 11 months ago, when we took our last two little nibbles and bought some more Meta.

But it's certainly not an unreasonable assumption. The revenue growth alone, even if it decelerates, should take care of that. The question will now be Meta's profit margin. Can it at least remain stable as they start to light up all of this new data center infrastructure?

For us personally here at CSI, we have a full position in Meta. We've had Meta in our portfolio for a long time now. Coming up on a decade and a half.