The Asymmetric AI Bet: Why Meta Is A Company For All Markets

$META is currently trading at $615 and a PE of 21.03x. META is underperforming the S&P 500 both YTD and for the past year. In fact, it is valued below at least HALF of the S&P 500 companies, which have an average PE of 26.07x.

All of this while their revenue is growing at an incredible 3Y CAGR of 23.7%. Although this method has flaws at face value, if we use Peter Lynch's PEG metric (Adjusted for future earnings growth), $META would pass with flying colors, with a multiple of 0.79x!

This brings us to a simple valuation question: Do you, as an investor, believe $META is a worse company than at least half of the S&P 500? I would easily say no.

$META currently has quarterly revenue of $64.60B, up 28% year over year. For a company valued at over $1 trillion, it is growing like a high-growth medium-cap stock.

Despite revenue growth, investors have not rewarded the stock with a higher multiple. In that regard, based on multiples, $META is the cheapest it has been in a while.

$META's free cash flow took a significant dip in the most recent quarter, a whopping -80.6% YoY, to −$1.16B. Normally, this falloff would warrant a complete overhaul of the thesis; however, in this unique market, I would argue that this decline in free cash flow is actually a GOOD thing for the company.

Let's observe the free cash flow below:

This sharp drop in FCF stems from $META's massive CapEx spending on its supercomputing ambitions. More specifically, this CapEx is mainly for data center buildout. In this quarter, $META racked up a CapEx value of $33.94B

The graph shows the substantial CapEx needed to power $META's future, which does eat into FCF. That is why I believe that, for $META in particular, it is better to look at Operating Cash Flow (OCF), as it shows the cash flow generated by the business itself, separate from the AI CapEx race.

The OCF story is completely different from the FCF story because it removes the burden of CapEx. Although it's not incredibly impressive because of stagnation, this largely stems from legal fees, one-time costs, and the burden of reality labs (More on this later).

I believe strong revenue growth will eventually win out in the long run, and the divergence will lead to higher OCF.

While $META will also suffer in the near term due to CapEx amortization and GPU/server depreciation, as long as its core business stays steady and it continues to grow revenue while keeping operating margins in check, it will still benefit heavily from current spending and debt.

Based on current trends, I see no reason to think revenue will slow, users will decline, or margins will contract significantly (Ignoring legal fees).

$META is an interesting case where EBITDA tells the whole story. Recent bills, such as the "Big Beautiful Bill" have actually increased $META's tax burden on income. As such, you can see EBITDA outpace Net Income here, which is why earnings may not tell the complete story.

Due to the heavy tax burden and legal volatility, $META's income is depressed despite revenue acceleration.

As expected, since $META is focusing so heavily on CapEx, they have not been repurchasing shares in recent quarters. The justification for this however, makes sense since we can see the results of the re-investment into the business.

$META has accelerated its ad business directly because of LLMs. LLMs have helped $META better identify what kind of content each user aligns with, and by optimizing its algorithm to bring specific content to users, it has successfully increased time spent per user across its family of apps.

$META has reported growing engagement in ad impressions, driven by increased engagement across its family of apps.

Alongside that, during the Q2 FY2026 earnings call, Mark Zuckerberg made a comment directly attributing LLMs for this increased engagement and thus increased revenue:

As we can see, the cost of revenue is increasing and accelerating in recent quarters, suggesting heavy compute/server involvement. Despite higher server costs associated with LLMs, revenue growth continues to keep up with the increase in cost of revenue.

This suggests that the CapEx and debt issued by $META are being put to good use... and based on projections by $META itself and analysts, we have many quarters of growth to look forward to.

However, it is not all sunshine and rainbows- this will be something to monitor closely in the future. If cost of revenue significantly outpaces revenue growth, it may be time to take a smaller position in this stock.

Q3 FY2026 guidance is about as expected, and $META continues its trend of accelerated revenue growth. With this in mind we can check when this acceleration occured and map it to when $META really started pushing LLMs into their platform.

Here we see that over the last 8 quarters (the last 2 years), the YoY CAGR is in the high 20s; in the early 2020s, $META had a YoY CAGR in the low 20s. This clearly shows acceleration, and whatever they are doing with AI is working.

$META was once just Facebook, but now it is a growing family of apps that Mark Zuckerberg himself describes as "A business in a box." Muse, META One, WhatsApp, Instagram, Threads, Messenger, Reality Labs, and, of course, Facebook see 3.5B+ daily users, creating a moat that feels like an impenetrable fortress.

For a long time, others have questioned this moat because social media and virality can be flimsy. I bought into this idea for a long time too, but as $META faced changing times with competitors like TikTok and Snapchat, I realized that for the average consumer, $META truly has something for everyone.

The risk of Social Media is that eventually people will grow out of it- $META does not have this problem for two reasons.

Mark Zuckerberg's initial goal was to connect the world, and in a way, he has. $META is a primary source of communication for many people worldwide. Children connect with their parents, friends connect with each other, and colleagues stay in touch. This makes it hard for anyone to want to leave $META's ecosystem.

What's even more crucial in my opinion, is how many small businesses rely on $META:

I couldn't write an analysis of $META without mentioning Muse. As a Software/Data Engineer, I have a lot of experience with frontier AI tools like Claude and Codex; however, I've always wondered whether any of these tools would ever break into the mainstream and be used by regular consumers. Sure, enterprises use them situationally for non-coding purposes, but for normal people to feel the impact of AI, these tools need to be accessible to everyone.

OpenClaw was the first time I thought this was possible, but it seemed to only appeal to people who went digging for it. Since then, Muse has taken that role solidly. It fits Zuckerberg's vision of a "business-in-a-box" platform because it lets regular users on Instagram or Facebook not only connect in messaging but also connect their calendars and their internet bills, shopping, finance, entertainment the list goes on.

Anthropic is supposedly close to hitting 100B in annual revenue from subscriptions alone. If Muse can approach even a quarter of that, it would be an additional 25B in raw revenue for $META. At that point, CapEx spend becomes meaningless because this trillion-dollar company would keep accelerating and hit revenue growth beyond +30%.

Now, of course, this is hypothetical, but based on early reactions, I don't think it's unreasonable to think Muse can accelerate $META's revenue. Here is a great interview of Zuckerberg speaking about it:

Right from the man himself. One tried-and-true way to win in business is to create something that saves others money. There are examples of Muse doing just that:

Ignoring AI altogether, $META still has additional growth avenues.

For starters, they've started incorporating marketplace ads on Instagram. This is a small thing, but it lets $META earn a small commission on purchases. Obviously, this is nowhere near as big as Muse, but they may be able to scalp a couple billion annually in revenue with virtually no cost.

WhatsApp is completely unknown territory. For Americans, it may seem like an unpopular product, but in Europe and much of the globe, it is the primary method of communication for businesses and regular users alike. Currently, WhatsApp has ~2.3B daily active users, and virtually no ads run on this platform. Showing ads to 2.3B people opens up another avenue for $META.

Meta also has the Meta One subscription, which reportedly already has 15M subscribers. Cheap to purchase, but when leveraged against Meta's enormous user base, this can also add a couple of billion to their revenue.

When a company has a user base as expansive as $META, there are many directions it can take to generate more cash.

All these things sound amazing on paper, but one should ask: why is the stock price so depressed? Rarely is the market completely incoherent, yet a seemingly near-perfect stock can stay stagnant or fall. That is not the case for $META, because $META has flaws that scare institutions.

Any seasoned investor can see the incredible revenue momentum at $META; however, they will ask: If revenue is increasing so fast, why aren't operating income and net income following the same trend? If we look at operating margin, it's not that impressive:

Quarterly margins are mostly stagnant, except in the most recent quarter, when they fell sharply to 20.1%, down from previous quarters.

$META explains this in a couple of ways. For one, they were hit with more-than-normal legal fees (to the tune of 2.4B) and severance fees from the layoffs. This, combined with their CapEx spend, caused amortization of their debt for the quarter. Another big loss leader for $META is the operating costs of Reality Labs- their primary hardware designer.

These headwinds cannot be overstated; however, most, if not all, are either self-inflicted (like Reality Labs) or expected to fade, like the legal troubles $META is facing.

I would only be concerned if these legal troubles significantly impact revenue.

We can review the debt structure to get a holistic view of the company.

A Monte Carlo simulation using a terminal earnings-multiple model that maps future revenue shows that, in over 98% of cases, the stock is currently undervalued.

The estimates I ran with for revenue growth were also extremely conservative at 20% (Below 5Y and 3Y CAGR). This is a clean single: $META might be undervalued right now, but what does a DCF say?

Even with $META's current compressed cash flow, the company is still set to grow significantly in the near future. This assumes $META continues to outperform the economy over the next 10 years, compounding its revenue growth.

10 years is a long time, but with a company as ingrained in so many people's lives as $META, this is a defense you can reasonably make.

These are the risks I think Wall Street and most investors worry about with $META. Like any stock, there is a world where these risks come to fruition, and $META ends up a dud. Here is what I'm looking for to decide whether or not to continue investing in $META:

It's understandable why some are fearful of $META right now. The company is doing a big transition away from what it's used to. $META can no longer be considered an asset-light business focused primarily on social media; instead, it is a major participant in the AI gold rush and operates a "business-in-a-box" model, leveraging its immense data & user moats to drive engagement and ad revenue across its platforms.

$META is seeing accelerating revenue growth, despite being depressed in terms of price- a clear discount in valuations that is too juicy to ignore. Backed by an impenetrable moat and a growing family of apps that is slowly integrating complex agents and delivering useful AI to everyday users, it seems like a no-brainer.

Despite some legal headwinds- which so far have been minimal in cost and fears of ever-growing debt to fund their compute- the fortress of cash $META has built over the years can safely cushion it from poor earnings and cash flow. The best part is that, at any time, if the market gets tough or ROI on investment isn't there, $META can cut its losses and go back to its tried-and-true core business that we know works. This is not true for Frontier Labs and puts $META in a clear position to win out no matter the outcome.