01 — Understanding the Business
Meta reports two main segments:
Family of Apps, which includes Facebook, Instagram, Messenger, WhatsApp and related services, and Reality Labs, which includes Meta's virtual and augmented reality hardware, software and smart glasses.
Despite all the attention around AI and the metaverse, Meta is still overwhelmingly an advertising company.
In Q2 2026, total revenue reached $60.8 billion, while advertising revenue was $59.4 billion. Family of Apps generated $60.4 billion, compared with just $431 million from Reality Labs.
The advertising engine continues to perform extremely well.
Family daily active people reached 3.60 billion, up 3% year over year. Ad impressions increased 14%, while the average price per ad increased 12%. Total revenue grew 28%.
This is important because Meta is no longer dependent on fast user growth. Its user base is already enormous.
Instead, growth increasingly comes from improving engagement, recommendations and monetization across the existing ecosystem.
AI is already helping the core business
One of the most important points from the latest earnings call is that Meta's AI investment is not purely a future bet.
The company is already using large language models to improve recommendations on Facebook and Instagram, understand content more deeply and improve advertising relevance.
Management reported double-digit growth in time spent on Instagram, while Facebook video time increased 9% globally. Meta is also increasingly using AI to improve ad matching and conversion performance.
That gives Meta a strong position in AI.
It already has the three things many AI companies still need to build:
distribution, data and monetization.
Meta can introduce an improvement to billions of users, measure the impact almost immediately and monetize the result through advertising.
That is a meaningful competitive advantage.
Where future growth could come from
The most interesting opportunity outside advertising is probably WhatsApp.
Family of Apps "other revenue" exceeded $1 billion in a quarter for the first time, growing 73% year over year, mainly because of paid messaging and subscriptions.
Meta is also scaling Business Agents across WhatsApp, Messenger and Instagram. More than 1 million businesses were already using them, with monetization expected through subscriptions and usage-based pricing.
Beyond that, Meta is developing AI APIs, productivity tools and enterprise services, while management has also discussed potentially selling compute directly.
I see these opportunities as meaningful optionality, but I would not build too much of them into a base-case valuation yet. The core advertising business remains what matters most today.
02 — Analyzing Financial Health
Meta's underlying financial performance remains very strong.
The complication is that the company is becoming much more capital intensive.
Q2 revenue grew 28% to $60.8 billion, while operating income was $18.8 billion.
The reported operating margin fell to 31%, but the quarter included $2.4 billion of legal charges and around $1.2 billion of severance expenses. Management said operating income would have increased year over year excluding those items.
So the core business remains highly profitable.
The real pressure is coming from AI investment.
Research and development expenses reached $21.7 billion in Q2, compared with $12.9 billion one year earlier. For the first half of 2026, R&D reached $39.4 billion.
Property and equipment also increased from $176.4 billion at the end of 2025 to $225.7 billion by June 2026.
Free cash flow is where the change is most visible
In Q2, Meta generated $31.9 billion of operating cash flow.
Free cash flow was only $784 million.
The reason is simple: the company is investing heavily in servers, data centers and network infrastructure.
Meta now expects $130–145 billion of capital expenditures in 2026.
This does not mean the business is weakening.
It means a much larger portion of the cash generated by Facebook, Instagram and WhatsApp is being reinvested into infrastructure.
That makes one question particularly important:
Will today's AI investment create enough future cash flow to justify the amount of capital being deployed?
The balance sheet remains strong.
At the end of Q2 Meta had approximately $90.3 billion in cash and marketable securities and $83.7 billion of long-term debt.
Debt is increasing, but I do not see a solvency problem. The more important issue is whether the returns generated by this investment cycle will be high enough.
Reality Labs remains another cost to watch.
During H1 2026 it generated only $833 million of revenue while producing an operating loss of around $8.6 billion.
Meta can easily afford those losses today, but they still represent significant capital being invested into a business with uncertain long-term returns.
03 — Perform DCF Valuation
For valuation, I used my existing ValueSages DCF.
The main assumptions are:
- 15% annual FCF growth
- 9.88% discount rate
- 10-year forecast period
- 10x terminal multiple
- 15% margin of safety
Under those assumptions, the model calculates an intrinsic value of approximately:
$519.92 per share
Applying a 15% margin of safety gives a target purchase price of approximately:
$441.93 per share
The stock price used in the analysis was around $549.90.
The important part is not whether the exact intrinsic value should be $520, $500 or something slightly different.
The important part is what the assumptions tell us.
A 15% annual free cash flow growth rate for ten years is already a fairly strong assumption for a company of Meta's size.
That growth is certainly possible if AI continues improving the existing advertising business and new revenue streams begin contributing meaningfully.
But at roughly $550 per share, the market is already pricing in much of that success.
In other words, the valuation does not leave a particularly large margin for execution mistakes.
04 — Assess Leadership & Capital Stewardship
Mark Zuckerberg's operating record deserves significant credit.
Meta successfully navigated the shift to mobile, acquired Instagram and WhatsApp, responded to TikTok through Reels, recovered from Apple's privacy changes and rebuilt margins after the 2022 downturn.
The current AI investment, however, is probably one of the biggest capital-allocation decisions in the company's history.
During the first half of 2026, Meta repurchased no shares.
Instead, capital is being directed toward infrastructure, while the company also issued additional debt.
The priority is obvious:
AI infrastructure comes first.
The reasoning is understandable.
Management says Meta is currently constrained by available compute and still sees many high-return uses for additional capacity inside the existing business.
If better models continue improving recommendations, engagement and advertising performance, those infrastructure investments could generate very attractive returns.
The risk is that the company is committing enormous amounts of capital before the long-term economics of consumer AI, enterprise agents and frontier models are fully proven.
That makes capital discipline increasingly important.
Meta's governance structure matters here as well. Zuckerberg maintains effective control through the dual-class share structure, giving him significant freedom to pursue long-term strategies.
That can be an advantage when management is right.
It also means outside shareholders have limited influence if spending eventually becomes excessive.
05 — Build Investment Thesis
The positive side of the case is clear.
Meta has an exceptional advertising franchise, billions of daily users, very strong network effects and one of the most advanced advertising systems in the world.
The company is still growing revenue close to 30% despite its scale.
AI is already improving the core business rather than simply representing a future promise.
WhatsApp remains under-monetized, Business Agents are starting to scale, and enterprise AI creates additional upside.
The financial strength of Family of Apps also gives Meta the ability to fund investments that very few companies could even consider.
The risks are equally clear.
Capital expenditures have increased dramatically.
Free cash flow is currently under pressure.
Debt is rising.
Reality Labs continues to lose billions.
And some of the largest AI investments may take years before their returns become visible.
The valuation adds another layer of risk.
At roughly $550 per share, Meta is not priced like a business where very little needs to go right.
The market already expects significant free cash flow growth and successful monetization of the current AI investment cycle.
That does not make META a bad investment.
It simply means the quality of the company and the attractiveness of the stock are not exactly the same thing.
The key relationship I will be watching over the next few years is:
Operating Cash Flow → AI Investment → Free Cash Flow
If Meta proves that today's extraordinary infrastructure spending can produce significantly higher future cash flow, the current investment cycle could create enormous shareholder value.
If capital spending remains permanently elevated without a corresponding increase in monetization, the current valuation becomes much harder to justify.
For me, that is the central question behind Meta today.