Meta Platforms

Meta agreed to pay $17 billion in fines and make platform-level safety changes, including daily use limits, notification blocking, and overnight restriction aimed at teen users in a settlement. The proposed deal required court approval.

Why it matters: The settlement, if approved, would lift a large legal overhang on Meta's stock, with prior reports of legal liabilities being materially higher than the proposed $17 billion in the settlement. We expect behavioral changes imposed on Meta to only marginally trim teen time spent on Meta's properties. Our view is informed by the fact that while the settlement imposes a two-hour limit on app time, the average time spent by teens on Meta's platforms in the US is less than 30 minutes. With a marginal reduction in time spent, we don't see a large revenue impact due to two factors. One, teens make up less than 10% of the overall US user base. Two, teens' revenue contribution is even lower. The real value in teen users is their lifetime value, which is maintained in this settlement.

The bottom line: We maintain our $850 per share fair value estimate for wide-moat Meta with the present value of the legal settlement of $17 billion, which is to be paid over 10 years, less than 1% of the firm's market cap. We continue to view Meta as materially undervalued. Other than its legal challenges, investors continue to be worried about Meta's returns on its artificial intelligence investments, which we view as the biggest overhang on the stock. Our view is that Meta has a multivariate AI monetization opportunity, with the firm monetizing via ads, selling tokens to external customers via Meta API, and selling excess compute to other AI labs. The market is highly underpricing the latter two, in our view.

Coming up: We expect Meta to formalize plans to sell excess compute to external AI labs in the next couple of quarters, leading to material revenue contributions in 2027 and beyond.

Our fair value estimate for Meta is $850 per share, implying a 2026 adjusted price/earnings multiple of 26 times and an enterprise value/adjusted EBITDA multiple of 13 times.

We forecast Meta's sales growing at a 22% compound annual growth rate for the next five years, spearheaded primarily by an increase in average revenue per user, with user growth also chipping in. We additionally expect Meta to leverage its AI data center buildout by renting out excess capacity to AI labs. We expect this business to grow from practically $0 in 2026 to $40 billion by 2030.

On advertising, we believe Meta has a strong monetization opportunity ahead of it in Asia and the rest of the world. While we expect advertising sales from North America and Europe to grow steadily, we believe increasingly affluent and growing middle classes in Asia, Africa, and the Middle East will allow Meta to improve its ad monetization in those regions, lifting its overall top line.

While we expect Reality Labs sales to grow at a double-digit rate over the next five years, we believe Meta’s advertising juggernaut will remain the primary driver of its business and intrinsic value over our explicit forecast.

On the profitability side, we remain impressed by Meta’s ability to drive efficiency across its operational footprint, with the firm’s 2025 operating margins of 41%, up from the 2022 nadir of 25%. Looking ahead, we believe that the firm’s profitability will deteriorate, with operating margins declining to 36% over our explicit five-year forecast as increased compensation and depreciation costs eat away at the strong top-line growth.

We believe Meta merits a wide economic moat rating due to the firm's intangible assets and the potent network effect around its Family of Apps business. While the firm's Reality Labs segment continues to hemorrhage cash, we believe Meta's FoA business' strong competitive advantages will likely allow the firm to generate returns in excess of its cost of capital over the next two decades.

Family of Apps

We assign a wide moat rating to Meta's Family of Apps, or FoA, business segment. We believe that the firm has built significant intangible assets, primarily via the customer data it collects and a potent network effect that has enabled Meta to be the most dominant social media platform in the world.

Meta's FoA segment includes revenue from its social media applications including Facebook, Instagram, WhatsApp, and Messenger. The firm's dominance in social media is evidenced by its four primary applications constituting four of the six most popular social media applications globally. Also, Meta's scale in the social media business is staggering. Almost 4 billion people use at least one of its applications every month. For context, according to various estimates, a little more than 5 billion people in the world have access to the internet, implying that around 75% of people connected to the internet globally are users of Meta's applications.

The vast majority of Meta's massive user base across its application ecosystem uses its applications free of charge. Instead of paying Meta a subscription fee to use its applications, they provide Meta with an audience that it can sell advertisements to. Meta can accumulate data on users, such as demographic information, likes/dislikes, and topics of interest to feed into its advertising engine, which lets advertisers target ads placed on Meta's properties.

We view Meta's collection, storage, and subsequent monetization of this rich, high-quality data as an intangible asset that provides the firm with a meaningful competitive advantage over other social media platforms and advertising venues. This data, and the ad technology and algorithms built on top of it, provide Meta a peerless position when it comes to monetizing its large user base. Further, as Meta's data collection and ad-targeting algorithms have gotten better, the firm has also flexed its pricing power. A prime example of this pricing power, and the increasingly strong value proposition the firm offers advertisers, is the rapid growth in Meta's average ad revenue per user, which has increased to more than $40 globally at the end of 2023 from around $25 in 2018.

Further, Meta's ad-targeting and content recommendation algorithms improve as more users give it their data by using its applications. This dynamic creates a potent network effect with the value of its application ecosystem increasing as more people use it. These network effects serve to create barriers to success for new social network upstarts, who would start off subscale, as well as barriers to exit for existing users who might leave behind friends, contacts, pictures, memories, and more by completely departing to alternative applications such as Snapchat or TikTok.

Meta has leveraged its scale and large installed base to create new avenues of user engagement that directly compete with applications that stand to threaten its social media dominance. When threatened by Snapchat, Meta launched Stories, a Snapchat copycat, that allows FoA users to post short picture or video collections that disappear after 24 hours. Similarly, when threatened by TikTok and its engaging short-form video content, Meta rolled out Reels, a short-form video offering nearly identical to TikTok.

While Stories and Reels can be seen as defensive maneuvers aimed to keep users on Meta’s applications, Meta has demonstrated its ability to use this copycat model as an offensive, land-grab strategy as well. When Twitter/X was undergoing substantial turmoil following Elon Musk's takeover of the platform, Meta launched Threads, a Twitter/X copycat. In all three cases, Stories, Reels, and Threads, Meta has been able to create strong, monetizable products that have hundreds of millions of users across them. We believe Meta’s ability to weaponize its scale as a competitive advantage makes the economic moat around its business more durable.

Meta's scale is also important when contextualizing the firm's ongoing investments in AI. Meta's capital expenditure, mostly on data centers equipped with expensive graphics processing units, has led investors to question whether the firm's investments in AI will yield meaningful results and generate returns in excess of the firm's cost of capital. When discerning the impact of these AI investments on Meta's return on invested capital profile and its durable competitive advantage and differentiation, we see positives and negatives.

On the positive side, the firm could leverage AI to improve its ad tech business, improving advertiser ROI and encouraging more ad dollars to be spent on its platform. Investments in AI that serve this end are, in our view, explicitly value-accretive. By investing within its moat, Meta could leverage its AI infrastructure to further pull away from competition by driving a higher ad revenue per user in its highly profitable ad business. However, investments to drive better ad targeting are nothing new. Also, the firm's AI investments could be leveraged to drive more engaging content for its end users on Facebook and Instagram. Again, such investments in better user engagement are ultimately value-accretive as they increase Meta's value to advertisers who can place ads for more engaged users.

On the other hand, the firm's investments in generative AI and its push to have Llama as the chatbot of choice for its users don't inspire the same confidence in us. First, the firm lacks any meaningful monetization strategy around its chatbot. We understand that if users have a competent chatbot within their Meta application of choice they don't need to leave Meta's properties, boosting time spent on Meta's application ecosystem. However, at the same time we'd imagine it'd be significantly cheaper for Meta to have a licensing agreement with a company like OpenAI/Anthropic/Perplexity, which could provide Meta with chatbots that would have the same end user-engagement result.

Also, we see the investments that public cloud vendors are making in AI as beneficial for public cloud infrastructure providers like Amazon, Microsoft, and Alphabet. Meta, on the other hand, does not have a presence in the public cloud market and isn't planning to become a public cloud vendor either. As a result, it strikes us as odd that the company would invest billions of dollars in developing a technology that doesn't have an obvious monetization strategy behind it. In sum, while we like Meta's investments in AI as they relate to the firm's core advertising business, we remain skeptical of the long-term value added by the firm’s investments in user-facing chatbots.

Reality Labs

We believe Meta's Reality Labs business merits a no moat rating. While the firm's investments in metaverse and virtual/augmented reality could lead to profitable growth in the future, the segment continues to burn capital for Meta, with the segment’s operating losses exceeding $16 billion in 2023. Due to its insignificant size, as a proportion of Meta's overall business, it does not preclude us from viewing the overall firm as having a wide moat.

Meta's core advertising business has benefited greatly through improved ad targeting and content recommendation algorithms as well as a secular increase in digital advertising spending.

Meta's scale, with the majority of the world's internet-connected users accessing its applications, allows it access to high-quality user data which it can package and sell to advertisers.

The firm has an opportunity to drive more ad inventory growth, leveraging new products such as Threads while also improving its monetization of ads on more nascent features such as Stories and Reels.

Meta's investments in Reality Labs and generative AI stand to lose the firm billions of dollars annually, taking some of the shine off its overall business.

The firm has a monopoly case against it in the US which could potentially force it to break up, severing some of the scale advantages it has built up over time.

Meta has disproportionately benefited from increased ad spending by Chinese retailers including Temu and Shein. A slowdown in spending by these firms could hit Meta's growth.