To The Moon

Takeaway

Meta’s core ad business is still strong. AI is already helping user engagement, ad delivery and conversion. The main debate is whether this can justify a 2026 capex budget of US$125–145 billion.

The bull case is clear: better recommendations, stronger ad tools and higher GPU use can support growth. The bear case is also valid: external AI services may face falling prices, high competition and weak returns.

For Q2, the key numbers are ad impressions, ad pricing, operating margin, capex and free cash flow.

$Meta Platforms, Inc.(META)$ will report Q2 2026 results after market close on 29 July. Market estimates point to revenue of about US$60.2 billion, close to the top end of management’s US$58–61 billion guidance.

Revenue growth matters, but investors are more focused on AI spending and cash returns.

Meta’s core business remains in good shape. In Q1, revenue reached US$56.3 billion, up 33% year on year. Ad revenue was about US$55 billion. Ad impressions rose 19%, while average ad price increased 12%. Family daily active people reached 3.56 billion.

Operating income was US$22.9 billion, with an operating margin of 41%.

These numbers show that Facebook, Instagram and WhatsApp are still growing in both engagement and monetisation.

AI is already supporting this growth. Better recommendation systems have increased time spent on Reels and video. Improved ad models have also lifted conversion rates. More advertisers are using Meta’s AI tools to create images, copy and other ad content.

This is the strongest part of Meta’s AI story. The company already has user data, ad demand, content supply and a large auction system. Any improvement in targeting or conversion can quickly turn into revenue.

Some estimates suggest that AI-driven ad improvements could add US$7–12 billion in annual revenue. With high incremental margins, this could support several billion dollars of extra profit.

This is why many investors remain positive. Meta does not need to build a large new AI business to benefit from AI. It can improve the economics of its existing ad platform.

The concern is the size of the investment.

Meta expects 2026 capex of US$125–145 billion, up sharply from US$72.2 billion in 2025. The midpoint is US$135 billion, almost 87% higher.

In Q1 alone, capex reached US$19.8 billion, while free cash flow was US$12.4 billion. This means cash spending is already rising faster than near-term cash generation.

Management says the money is going into data centres, chips, servers and other AI infrastructure. Some of this spending is for future demand, so the full return may take years.

The bear view is that Meta is investing too much in a market where model and token prices are falling. Developers can switch between OpenAI, Anthropic, Google and open-source models. This limits pricing power.

Meta is also exploring cloud hosting and compute rental. A possible deal with Anthropic could be worth up to US$10 billion over two years. This may improve GPU use, but an annualised US$5 billion of revenue would still be small compared with annual capex of around US$135 billion.

Supporters see this as a new growth path. Meta may use spare capacity, enter enterprise AI and build a more complete platform around models, agents and cloud services.

Sceptics see a different risk. The move may suggest that internal demand is not enough to absorb all planned capacity. If Meta has to rent out GPUs at lower prices, returns may stay weak.

Investors should also watch long-term leases and purchase commitments. These are not the same as normal bank debt, but they still lock in future cash payments. As depreciation, leases and stock-based compensation rise, reported profit may look stronger than real cash generation.

Valuation is also mixed. Meta’s forward earnings multiple is lower than some large tech peers. Its free cash flow multiple is less attractive because capex is already very high.

This means the stock can still perform if ad growth stays strong and capex starts to peak. Upside may be limited if spending keeps rising without a clear improvement in free cash flow.

For Q2, I would focus on five items: ad impressions, average ad price, operating margin, quarterly capex and free cash flow.

The earnings call should also answer three questions. How much is AI improving advertiser returns? How will Meta price its cloud and compute services? Will management raise capex again?

Meta’s ad business is strong enough to support the AI buildout for now. The next step is proving that the spending can create durable cash returns.