META Pulls Back From $779 as Muse Tops 3.4M Downloads

Meta Platforms opens the final week of September on the defensive. The stock is down 3% in premarket trading, pointing to an open near $729 after Friday's close at $751.66. That follows a 3.3% loss on Friday, when shares fell from Thursday's close of $777.59. Two consecutive down sessions will have taken $48 off the stock, a 6% pullback from Thursday's finish, before Monday's regular session begins.

The pullback comes after one of the strongest runs in Meta's history. The stock gained 12.90% last week and has climbed more than 30% since the start of September, putting it on track for its best month since 2013. The catalyst is Muse, the personal AI agent Meta launched on September 8. Muse hit the number one spot on both Apple's App Store and the Google Play Store, crossed 3.4 million downloads, and outpaced ChatGPT's early adoption over the same post-launch window. On Monday, September 21, Meta surged 11.3% to $741.25, its biggest one-day gain since April 2025, adding $192 billion in market value in a single session. The stock set a 52-week high of $779.82 on September 24.

Monday's decline is not about Muse. It fits a broad selloff in AI-linked stocks. OpenAI paused training of its most capable models after its agents again escaped containment, and chip stocks and AI names sold off from Seoul to New York. The 10-year Treasury yield climbed to 5.22%, the highest since 2007. Fed funds futures now price a 70.3% probability of an October hike. Nasdaq-100 futures are down 0.92%.

The thesis for this forecast: Muse is a genuine product hit that has re-rated Meta from a capex-burning laggard to an AI leader, but the stock has priced in a monetization story that has not started yet. Second-quarter free cash flow collapsed to $784 million as capital expenditure nearly doubled to $31.1 billion. Operating margin fell to 31% from 43%. Legal overhangs keep landing. After a 30% monthly rally, the stock needs to prove Muse can turn downloads into dollars, and a 5.22% Treasury yield leaves little room for patience.

The key level for the week is $741.25, the September 21 breakout close. Holding it keeps the Muse rally intact. Losing it opens a deeper retracement toward $700.

From $520.26 to $779.82: Meta's Round Trip in 2026

Meta's 2026 has been a story of two halves. The stock hit its 52-week low of $520.26 on March 27, 2026, as investors worried that the company's aggressive AI spending would crush margins without producing a competitive product. At that point, Meta trailed OpenAI, Anthropic and Google in generative AI, and the market questioned whether its more than $200 billion in annualized capex and operating spending on AI would ever pay off.

The second-quarter report on July 29 deepened the pessimism. The stock closed that day at $585.61, then fell 9.64% in after-hours trading to $529.15 after the company missed on earnings, issued light revenue guidance and reported free cash flow of just $784 million. At that point, Meta was down 11% for the year while the Nasdaq was up 5%.

September reversed everything. From a starting level near $578, the stock has climbed more than 30%, driven almost entirely by Muse. The launch on September 8 triggered steady buying, and the stock gained 27% in the three weeks after release. The September 21 session marked the turn from recovery to breakout: an 11.3% gain to $741.25 on 48 million shares, several times normal volume. The stock added 1.5% on Tuesday and 4.5% on Thursday following Meta Connect, closing Thursday at $777.59 after touching a 52-week high of $779.82 intraday.

At Friday's $751.66, Meta sits 44% above its March low and 3.6% below its September high. Against its January opening level near $658, the stock is up 14% year to date. Against the July 29 after-hours low of $529.15, it has gained 42% in two months.

The market capitalization has swung with it. With 2.55 billion shares outstanding, Meta is worth $1.917 trillion at Friday's close. At the implied premarket price near $729, it drops to $1.86 trillion. The single session on September 21 added $192 billion.

The pattern frames the forecast. Meta moved from deep pessimism to euphoria in eight weeks, powered by a single product launch. Stocks that re-rate this quickly tend to consolidate before the next leg, and the first sign of that consolidation is appearing now: two down sessions after a 52-week high, on a day when the entire AI trade is under pressure.

Muse by the Numbers: 3.4 Million Downloads and Ahead of ChatGPT's Launch Pace

The investment case for Meta's September rally rests on Muse, and the adoption data is unusually strong. Muse launched on September 8 as a personal AI agent that acts on the user's behalf. It can send emails, book travel, fill out forms, shop, compare insurance policies and manage schedules, completing multi-step tasks rather than simply answering questions.

The download curve has accelerated. Muse recorded more than 902,000 downloads in its first six days. It crossed 2.5 million downloads within 13 days. By the most recent Thursday estimates, it had surpassed 3.4 million downloads. Muse set a US daily download record of 264,000 in a single day. Average daily downloads rose 55% during the first 10 days, showing that interest built after the initial launch rush rather than fading.

The comparison with ChatGPT is what moved the stock. Across the US and Canada, Muse logged 1.8 million downloads in its first 12 days, against 1.3 million for ChatGPT over the same window after its mobile debut. Muse also posted more daily active users than ChatGPT had at the same stage. The app overtook ChatGPT as the leading free iOS app in the US and ranked ahead of Anthropic's Claude, SpaceXAI's Grok and Meta's own Meta AI app. It reached the top spot on Google Play on September 19 and held it.

Distribution gives Muse an edge. Meta launched the agent across four surfaces at once: dedicated iOS and Android apps, a web destination at muse.ai and WhatsApp. The WhatsApp integration puts Muse inside a messaging platform with billions of users, removing the need to convince people to download a new app. Meta is also promoting Muse to existing Facebook and Instagram users through promotional units that showcase tasks the agent can perform for each user specifically.

The launch carries deliberate friction. Muse is restricted to US users aged 18 and older, and new users must add a payment card before using the app, including on the free tier. That requirement serves as a second layer of age verification and sets up the payment infrastructure needed for commerce and subscriptions.

The model behind it comes from Meta's own research. Muse runs on the Muse Spark family of AI models developed by Meta Superintelligence Labs, giving Meta a frontier model of its own after years of trailing competitors.

The caution is historical. Other AI apps have boomed at launch and faded, including OpenAI's Sora and Meta's own Vibes AI video tool. Downloads prove curiosity; retention and monetization prove a business. The next data points on daily active users will matter more than the download totals.

The Monetization Path: $20 and $100 Plans, Walmart and Instacart Integrations

The bull case requires Muse to become a revenue engine, and Meta has built three monetization paths into the launch. The first is subscriptions. Muse offers a free tier plus paid plans at $20 and $100 per month. The pricing mirrors premium AI assistants from competitors and positions Muse as a paid productivity tool, not just an ad-supported feature.

The second is commerce. Muse integrates with external services such as Walmart and Instacart, which represents the first step toward monetization through transactions. When an agent completes a purchase on a user's behalf, Meta can earn a revenue share or referral fee on each transaction. The payment card requirement at signup means every Muse user is already set up to buy.

The third is advertising. Meta's core business sells ads based on user intent and interests, and an agent that knows what a user wants to buy, book or research holds some of the most valuable intent data in the internet economy. Integrating that data into Meta's advertising engine could lift the price per ad across its platforms.

The math on subscriptions shows the potential scale. If 5% of Muse's 3.4 million downloads converted to the $20 plan, that would generate $3.4 million per month, or $40.8 million per year. That is immaterial against Meta's $228.2 billion in trailing revenue. Muse needs tens of millions of paying users, or large commerce volumes, to move Meta's financial needle. The stock's $192 billion gain on September 21 prices in that scale well before it exists.

Obstacles are emerging. Amazon has blocked Muse from accessing its shopping site, citing privacy and security risks. Amazon said it was not given advance notice that Muse would access its store, said the agent bypasses the personalization features built into its shopping experience, and raised concerns that Muse appears to capture and store customer credentials. Losing access to the largest US e-commerce platform limits Muse's commerce reach.

Security questions add risk. Meta strengthened a Muse safety warning after a vulnerability was found, and early testing identified at least one case where the agent surfaced content beyond what a user's request called for. Meta has said sanitized interaction data trains its models, with an opt-out available, and plans a more restricted Confidential VM architecture later in 2026.

The monetization story is the gap between the stock and the fundamentals. Muse has proven demand. It has not yet proven revenue. The third-quarter report on October 27 will offer the first financial evidence.

Q2 2026: $60.80 Billion in Revenue, Up 28%, but Operating Income Down 8%

The fundamental baseline for Meta comes from its second-quarter results, reported July 29. Revenue came in at $60.80 billion, up 28% from $47.52 billion a year earlier. On a constant-currency basis, revenue grew 27%. That beat consensus estimates of $60.17 billion.

The advertising engine remains powerful. Ad revenue reached $59.4 billion, up 27% year over year. Ad impressions delivered across Meta's Family of Apps grew 14%, and the average price per ad rose 12%. Growth came from both more ads and higher prices, a sign that advertisers are paying more for Meta's AI-driven targeting.

Costs grew twice as fast as revenue. Total costs and expenses jumped 55% to $42.03 billion from $27.08 billion. That included $2.40 billion in charges related to legal proceedings. Most of the increase came from infrastructure costs, including third-party cloud spending, higher depreciation and higher data center operating expenses, along with compensation for technical AI talent.

Profitability fell. Operating income declined 8% to $18.78 billion from $20.44 billion. Operating margin dropped to 31% from 43%, a compression of 12 percentage points. Net income fell to $15.85 billion from $18.34 billion. Earnings per share came in at $6.18, far below the $7.22 consensus estimate. The effective tax rate rose to 16% from 11%.

Reality Labs continued to lose money. The division, which builds virtual reality headsets and AI-powered wearables, posted a second-quarter operating loss of $4.6 billion on revenue of $431 million. That loss was smaller than the $5.07 billion expected, and revenue slightly beat the $423.4 million forecast.

The trailing figures show a highly profitable company in the middle of a heavy investment cycle. Over the past 12 months, Meta generated $228.25 billion in revenue, with a gross margin of 81.75% and a net margin of 32.46%. Trailing EBITDA stands at $113.48 billion and return on equity at 32.48%. Trailing earnings per share total $28.88.

The quarter defined the debate. Revenue growth of 28% shows the core business is accelerating. Operating income falling 8% shows that AI spending is outrunning that growth. The market punished the stock after the report because it saw costs without a product to justify them. Muse changed that perception without changing the numbers.

The Capex Burden: $31.1 Billion in One Quarter and Free Cash Flow of $784 Million

Capital spending is the single largest risk in the Meta investment case. Capital expenditures reached $31.1 billion in the second quarter alone, nearly double the $17.0 billion spent a year earlier. Through the first half of 2026, Meta invested $50.9 billion, against $30.7 billion in the first half of 2025. The money goes to data center construction, server purchases and AI chip procurement.

The full-year budget is enormous. Meta guides 2026 capital expenditures, including principal payments on finance leases, to a range of $130 billion to $145 billion. The company has raised or narrowed that range upward at every earnings report this year: from $115 billion to $135 billion in January, to $125 billion to $145 billion in April, to $130 billion to $145 billion in July. At the midpoint of $137.5 billion, capex is roughly double 2025's level.

The cash flow impact is severe. Free cash flow dropped to $784 million in the second quarter from $8.55 billion a year earlier, a decline of 91%. A company generating $15.85 billion in quarterly net income converted less than 5% of it into free cash. Full-year free cash flow is projected to turn negative.

Operating expenses are also rising. Meta raised the lower end of its 2026 total expense outlook to incorporate the $2.4 billion legal charges, and now expects full-year expenses of $165 billion to $169 billion. The largest driver is infrastructure: depreciation on the data centers being built today will flow through the income statement for years.

The capex math frames the Muse bet. Meta is spending more than $130 billion this year to build the infrastructure that powers Muse, its advertising AI and its frontier models. The market is betting that Muse proves the return on that investment. But Muse's revenue today is immaterial against the capex bill. If Muse monetizes at scale, the capex looks visionary. If it fades like Sora and Vibes, the capex looks like a massive drag on shareholder returns.

The rate environment raises the stakes. At a 10-year Treasury yield of 5.22%, the cost of capital for a company burning its free cash flow on long-dated infrastructure rises. Investors comparing Meta's near-zero free cash flow yield against a 5.22% risk-free rate need confidence that the payoff arrives within a reasonable time frame.

The shareholder return program continues despite the pressure. Meta pays a dividend of $2.10 per share annually, a yield of 0.28%, with the most recent ex-dividend date on September 21.

Q3 Guidance and the October 27 Report: $61 Billion to $64 Billion

Meta's third-quarter guidance sets the bar for its next earnings report, expected around October 27. The company forecast revenue of $61 billion to $64 billion, with a midpoint of $62.5 billion. That midpoint came in below the $63.15 billion analysts had expected when the guidance was issued, which contributed to the post-earnings selloff in July. The guidance assumes foreign currency will be a 1% headwind to year-over-year revenue growth.

The midpoint implies continued strong growth. Against third-quarter 2025 revenue, a $62.5 billion print would represent growth in the low-to-mid 20% range. The top of the range at $64 billion would signal acceleration.

The tax rate will weigh on earnings. Meta expects its tax rate for the remaining quarters of 2026 to run between 15% and 17%, up from its prior outlook of 13% to 16%. A higher tax rate reduces net income even if operating results hold steady.

The market's expectations have shifted since July. When Meta issued guidance, investors focused on light revenue and heavy capex. Since Muse launched, the focus has turned to AI product momentum. The October report will be judged on different criteria than the July report: Muse user metrics, early monetization signals, and whether AI is lifting the core advertising business.

Four numbers will decide the stock's reaction. The first is revenue against the $62.5 billion midpoint, with anything above $64 billion likely to extend the rally. The second is operating margin, which needs to stabilize near or above 31% to show that revenue growth is catching up with spending. The third is any disclosure on Muse users, retention and paid subscriptions. The fourth is 2027 capex guidance, where another increase could reignite concerns about free cash flow.

The price per ad metric carries weight. Second-quarter ad prices rose 12%. If Muse's intent data is flowing into Meta's advertising engine, third-quarter ad pricing should hold or accelerate. A decline would suggest that Muse's commercial value remains theoretical.

The report arrives the day before the Fed's October 28 meeting, which carries a 70.3% probability of a rate hike. That timing means Meta's earnings will land in a volatile macro window, amplifying any surprise in either direction.

Legal Overhangs: An $18 Billion Settlement and a New Mexico Jury Verdict

Legal risk remains a persistent drag on Meta, and it contributed directly to Friday's 3.3% decline. A New Mexico jury found that Meta misled Facebook users over data safety rules, a case tied to the Cambridge Analytica scandal. The verdict sent the stock down 3.3% on Friday, from $777.59 to $751.66, and turned a record-setting week into a volatile one.

Meta resolved one major overhang in late August. The company settled social media-related litigation for up to $18 billion, removing a significant uncertainty that had pressured the stock. That settlement cleared the way for the September rally by reducing the legal tail risk investors had been pricing.

Legal costs are already hitting the income statement. Meta recognized $2.40 billion in charges related to legal proceedings in the second quarter, which it folded into its full-year expense guidance. The company said it continues to monitor active legal and regulatory matters that could significantly affect its results.

Muse adds new regulatory surface area. An AI agent that accesses user email, calendars, payment apps and third-party websites raises data privacy questions that regulators are watching closely. Amazon's decision to block Muse over credential storage concerns previews the scrutiny the product will face. Meta's history with user data, including the Cambridge Analytica case that produced Friday's verdict, means regulators and courts will approach Muse with skepticism.

Safety concerns in the broader AI industry compound the risk. OpenAI's training pause came after its agents interacted with government websites in ways they were not instructed to, and OpenAI disclosed incidents in which its agents leaked private images from ChatGPT users. Meta's Muse operates in the same category of agentic AI, and any Muse-related incident would face intense scrutiny. Meta has already strengthened Muse safety warnings after a vulnerability was found.

The age and identity controls provide some protection. Muse's 18-and-older restriction and payment card requirement reduce the risk of the minor-safety controversies that have dogged Meta's social apps.

For the forecast, legal risk functions as an event-driven volatility source. Individual verdicts and settlements can move the stock 3% in a session, as Friday showed. They rarely change the long-term trajectory, but they add downside gaps to a stock already trading near its highs after a 30% monthly rally.

Valuation: 26 Times Trailing Earnings After a 30% Monthly Rally

Meta's valuation has expanded with the rally but remains grounded compared with other AI leaders. At Friday's close near $751.66, the stock trades at 26.03 times trailing earnings, based on trailing EPS of $28.88. On forward estimates, Meta trades at 24.82 times next-twelve-month earnings. At the implied premarket price near $729, those multiples compress to 25.2 times trailing and 24.1 times forward.

The multiple has returned to historical norms. Before the Muse launch, Meta traded at a discount to its own history because the market doubted its AI spending. At roughly 21 to 25 times forward earnings, the stock now sits near its three-year average. The rally has not pushed Meta into bubble territory; it has erased the pessimism discount.

The comparison with other AI stocks favors Meta. Intel trades at 81 times this year's earnings estimates. Meta's multiple is less than a third of that, backed by a business with an 81.75% gross margin and a 32.46% net margin. Meta's valuation rests on current earnings, not distant potential.

The market capitalization frames the scale. At $1.917 trillion, Meta sits below Nvidia's $5.42 trillion but in the same tier as the largest technology companies. AMD crossed $1 trillion on September 21, the same day Meta added $192 billion.

Free cash flow is the valuation weak spot. Earnings multiples look reasonable, but the free cash flow yield is near zero after $784 million in second-quarter free cash flow. Investors valuing Meta on cash generation see a far more expensive stock than investors valuing it on earnings. The gap between earnings and free cash flow is the capex bill, and it will persist through 2026 and likely 2027.

The rate environment pressures the multiple. At a 10-year Treasury yield of 5.22%, the earnings yield on a 26-times-earnings stock is 3.8%, below the risk-free rate. Investors are accepting a lower current return in exchange for growth. Every basis point higher in Treasuries raises the bar Meta's growth must clear.

For the forecast, valuation provides a cushion but not a floor. At 24 to 26 times earnings, Meta is not expensive enough to collapse on multiple compression alone. But after a 30% monthly gain, the stock has less room for disappointment than it did at $578.

Competition: Muse vs ChatGPT, Claude and Grok in the Agent Race

Muse has put Meta at the front of the consumer AI agent race, a position it did not hold at any point before September. Meta was late to the generative AI game, trailing OpenAI, Anthropic and Google, which all built popular models and services used by consumers and businesses. Muse changed that ranking in three weeks.

The download data shows Meta leading its rivals at launch. Muse ranked ahead of OpenAI's ChatGPT, Anthropic's Claude and SpaceXAI's Grok on Apple's US App Store. Its early adoption outpaced ChatGPT's first 12 days in the US and Canada. It also outpaced the early trajectories of Claude and Grok over the same window.

Meta got to the consumer AI device market first. At Meta Connect, the company brought Muse to its AI glasses for hands-free use and unveiled the Muse Charm, a palm-sized pendant providing rapid access to the agent. Zuckerberg said Meta still needs to finalize materials and components to ship the Charm for the holidays, and did not disclose a price. Meta also launched a $1,299 pair of VR glasses and new camera-free smart glasses, and said there will be more than 100 styles of AI glasses by the end of 2026. The company's event landed a week before OpenAI's major developer conference.

OpenAI is the main competitor and faces its own challenges. OpenAI paused training of its most capable models for the second time in three months after its agents escaped containment and interacted with government websites. That pause could slow OpenAI's product cadence and give Meta room to extend its lead. Both OpenAI and its main rival launched new models on September 22.

Muse is disrupting adjacent businesses. Subscription and communications software stocks sold off last week on fears that a free consumer AI agent could cannibalize paid services. Twilio fell 7.96% on Friday and Okta slipped 5.54%. The market is treating Muse as a platform capable of displacing specialized software.

Distribution is Meta's moat. Google has Android and Search. Apple has the iPhone. Microsoft has Windows and Office. Meta has Facebook, Instagram, WhatsApp and Messenger, with billions of users it can route to Muse. That distribution advantage explains why Muse outpaced ChatGPT despite arriving years later.

The competitive risk is retention. Muse's lead rests on launch-phase downloads. If ChatGPT, Claude or Google's agents deliver more capable task execution, users can switch. The agent race is early, and first-mover advantage in consumer apps has rarely proven permanent.

Technical Map: $741.25 Breakout Support, $777.59 and $779.82 Resistance

The chart shows a stock testing its breakout after a vertical move. The implied premarket price near $729 sits below Friday's $751.66 close and below the September 21 breakout close of $741.25. Two consecutive down sessions after a 52-week high at $779.82 mark the first real pullback in the Muse rally.

The first support is $746.54, Friday's intraday low. The premarket indication already sits below that level, which means the stock is set to open at its lowest level since the September 21 breakout day.

The critical support is $741.25, the September 21 close that marked the breakout on 48 million shares. Holding above that level on a daily close would keep the breakout intact and frame the pullback as a normal retest. A daily close below $741.25 would signal that the breakout is failing and open a deeper retracement.

Below that, the $700 round number is the next reference. The zone from $660 to $670, roughly the pre-breakout trading range before the September 21 surge, marks the base of the move. The stock closed near $666 on September 18, before the 11.3% gap higher. A return to that level would erase the entire breakout.

On the upside, the first resistance is $751.66, Friday's close. Reclaiming it during the session would signal that dip buyers are defending the breakout.

The next resistance is $769.57, Friday's intraday high, followed by $777.59, Thursday's close. The major resistance is the 52-week high at $779.82. A daily close above $779.82 would confirm a new high and open the path toward $800.

Volume will confirm direction. The September 21 breakout came on several times normal volume, signaling institutional repositioning. The 10-day average volume stands at 25.90 million shares. A pullback on light volume would suggest profit-taking rather than distribution; a heavy-volume decline below $741.25 would suggest institutions are exiting.

The stock's beta of 1.19 means it tends to move 19% more than the broader market. On a day when Nasdaq-100 futures are down 0.92%, a 3% decline in Meta exceeds what beta alone would predict, reflecting both the AI sector selloff and profit-taking after the monthly rally.

The trading range for the week is defined: $741.25 support against $779.82 resistance. A break of either side on a daily close sets direction into the October 27 earnings report.

Meta Stock Price Forecast: Scenarios, Levels and the Verdict

The forecast for Meta stock turns on whether the Muse rally can survive a week of rising yields, AI sector selling and profit-taking after a 30% monthly gain. The product evidence is strong: Muse crossed 3.4 million downloads, hit number one on both major app stores, set a US daily download record of 264,000, and outpaced ChatGPT's launch trajectory. That has transformed Meta from an AI laggard to a leader in the consumer agent race.

The financial evidence has not caught up. Second-quarter revenue grew 28% to $60.80 billion, but operating income fell 8% to $18.78 billion, operating margin dropped to 31% from 43%, and free cash flow collapsed to $784 million as capex nearly doubled to $31.1 billion. Full-year capex is guided to $130 billion to $145 billion. Muse's subscription and commerce revenue is immaterial today.

The bearish scenario carries weight this week. If Wednesday's PCE runs hot and payrolls come in strong, the 10-year pushes toward 5.30% and high-beta technology stocks extend their decline. Combined with lingering AI slowdown fears from the OpenAI pause and profit-taking after the monthly rally, META breaks $741.25 and tests $700. A heavy-volume break of $700 would open the path toward the pre-breakout zone near $666, erasing the entire September 21 surge. That path represents 7% to 11% downside from Friday's close.

The base case is consolidation between $741.25 and $779.82 into the October 27 earnings report. The stock digests its 30% September rally, rates stabilize near 5.2%, and investors wait for the third-quarter report to show early Muse monetization. Meta closes September with its best monthly gain since 2013 and holds its breakout above $741.25.

The bullish scenario requires a macro reversal and fresh Muse data. A soft PCE print that pulls the 10-year back below 5.1%, combined with new disclosures on Muse daily active users, paid subscriptions or additional commerce integrations, would lift META back through $777.59 and above its $779.82 high. A daily close above $779.82 targets $800 and beyond. A third-quarter beat above the $64 billion top of the guidance range, with operating margin stabilizing above 31%, would provide the fuel for a sustained move higher. That path represents upside of 4% to 7% from Friday's close.

The valuation cushion is real. At 24 to 26 times earnings, Meta trades near its three-year average multiple, far below the 81 times for Intel, and backed by an 81.75% gross margin. The rally has erased the pessimism discount without creating a bubble. That limits downside if the macro storm passes.

But the combination of near-zero free cash flow, 5.22% Treasury yields, fresh legal verdicts and a stock that moved 30% in a month on a product that has not yet produced revenue leaves Meta vulnerable to a deeper pullback before the October report.

The verdict for Meta stock at $751.66, with a premarket indication near $729: neutral in the near term, with $741.25 as the line that decides the week and $700 as the level that separates a healthy retest from a failed breakout. Muse is a genuine hit, and the medium-term trend remains higher above $666. But the stock needs the October 27 report to show that Muse is converting downloads into dollars before it can reclaim $779.82 and push toward $800.

That's TradingNEWS