Facebook, Instagram and WhatsApp-owner Meta Platforms (NASDAQ:META) has given UK retail investors a mixed but broadly encouraging read-through on the AI trade. Q3 revenue comfortably beat expectations, advertising remained strong and Meta’s new Muse AI agent is gaining users. However, the cost of building the AI infrastructure is becoming increasingly visible in margins and cash flow.
| Meta Platforms (NASDAQ:META) | Price: $689.65 (-1%) | Market cap: ~$1.76tn |
The immediate market reaction was positive: Meta shares surged nearly 7% in after-hours trading to ~$745, after closing the regular session at $697.43, but those early gains did not stick as investors mulled the still hefty cost of AI growth and the impact on cash flows.
As with SpaceX (NASDAQ:SPCX) 24-hours earlier, the 2026 investor narrative has clearly changed from buy AI, to buy superior AI returns.
The AI trade: What SpaceX and Super Micro earnings signal to investors
Meta Q3: reported vs expectations
| $bn unless stated | Q3 2026 reported | Wall Street expected | Beat/miss |
| Revenue | 64.60 | 63.37 | +1.23bn |
| Adjusted EPS | $7.40 | ~$6.5 | Beat |
| GAAP EPS | $3.83 | $6.54 | Miss* |
| Q4 revenue guidance | $73–76bn | $73.73bn | Midpoint above |
| Q3 capex | 34.9 | — | Up sharply |
| Q3 free cash flow | -$2.12bn | — | Major deterioration |
*GAAP EPS was depressed by a $9.89bn settlement charge. Excluding that charge, EPS was $7.40. Revenue increased 26% year-on-year to $64.6bn.
The Q4 outlook is particularly important. Meta expects revenue of $73bn-$76bn, versus the $73.73bn Wall Street consensus, suggesting that advertising demand remains resilient heading into the crucial Christmas quarter.
The AI trade: encouraging, but expensive
The most significant message from the results is that Meta is moving from AI infrastructure investment towards AI monetisation.
Meta spent $34.9bn on capital expenditure in Q3, up from $31.1bn in Q2 and $19.4bn a year earlier. It also lifted the bottom of its 2026 capex range from $130bn to $135bn, leaving guidance at $135bn-$145bn.
That spending is already hitting cash generation. Free cash flow swung to -$2.1bn, from positive $10.6bn a year earlier.
For investors, this is the critical tension in the AI trade: Meta has one of the world’s most profitable advertising businesses funding one of the world’s biggest AI infrastructure programmes, but the return on that investment is not yet fully visible.
The encouraging sign is Muse. More than 10 million users in the US and Canada had set up Muse accounts within six weeks, according to CEO Mark Zuckerberg. Management expects Meta’s AI businesses to start contributing incremental revenue in 2027.
Zuckerberg also said Meta expects Muse eventually to serve ‘billions of people’ and potentially generate more revenue per user than its existing products.
What analysts are saying
The tone from analysts has shifted noticeably.
Jesse Cohen, senior analyst at Investing.com, said: ‘The market has shifted from punishing Meta for spending to rewarding it for having a compelling product story.’
That captures the key change in investor sentiment. Earlier in 2026, higher AI spending frequently triggered share-price weakness because investors questioned whether the returns would justify the enormous investment.
Meta’s latest numbers suggest investors are increasingly prepared to give Zuckerberg the benefit of the doubt — provided revenue growth remains strong.
There are still important risks. Emarketer senior analyst Minda Smiley warned that Meta’s $25bn New Mexico privacy penalty highlights continuing legal risks and said privacy and safety concerns could intensify if AI agents become mainstream.
The bull case: Meta turns its enormous user base, advertising data and AI models into a dominant consumer AI ecosystem while advertising continues funding the build-out.
The bear case: AI infrastructure spending continues escalating faster than monetisation, depressing margins and free cash flow while regulatory and privacy risks increase.
Sharesify investor verdict
For UK retail investors, Meta’s results are more bullish for the AI trade than bearish.
The core advertising engine remains powerful enough to fund huge AI investment, revenue growth remains above 25%, Q4 guidance beat expectations and Muse provides an early indication that Meta may eventually turn AI from a cost centre into a significant new revenue stream.
But the numbers also show why investors should not simply equate higher AI spending with higher returns. Capex is approaching $145bn, free cash flow has turned negative, and management does not expect meaningful incremental AI revenue until 2027.
Bottom line: Meta’s latest earnings suggest the AI trade is moving into its next phase: from ‘who is spending the most?’ to ‘who can demonstrate the best return on AI investment?’ Meta has not proved that return yet — but its strong advertising growth and early Muse adoption give investors more evidence that the spending could eventually pay off.
You might also like:
