Meta Just Crashed 10% — Zuckerberg Has a $145B Problem

Meta has just fallen by around 10% after earnings, pushing one of the world's most profitable businesses back towards its 52- week low. But this was not because Meta's advertising business suddenly collapsed. We can see in the pre-market after reporting yesterday, it's trading around $535, currently down around 9%, although we expect this to fluctuate a lot. Now, in terms of where it sits, 52- week low isn't going to be that far off at 520. So, it leaves investors today facing a very important decision. Is this finally the meta buying opportunity that they've been waiting for? Or in fact, has Zuckerberg begun another spending cycle capable of destroying Meta's free cash flow for years? Because look at this heat map. It's showing us the pre-market moves while Microsoft's up around 9% yet Meta the opposite reaction. That's down around 8 1/2%. So this is not simply looking like a bad day for technology stocks. This is the market overall punishing Meta. I mean if we look from a year-to-ate perspective while Meta itself is standing around 19% for the year despite revenue growing at one of its fastest rates in years. And that contradiction is exactly what we need to investigate today. So in today's episode, I'm going to break down what actually went wrong. why Meta's advertising business may be considerably stronger than the share price reaction suggests and whether the company can ever earn an adequate return on its enormous AI investment. And then I'm also going to update my valuation model using three growth scenarios and show the exact assumptions behind the estimates because Meta's already dramatically underperformed the S&P 500 and NASDAQ before this latest collapse and that underperformance is now going to become very extreme. So, let's begin with the headline figures. Meta reported revenue of 60.8 billion, slightly above expectations and up an outstanding 28% year-over-year, but earnings per share came in 618 versus expectations of 722. That's a miss of around 14%. And the problem becomes obvious when we look below the revenue line. While revenue increased by around 28%, costs and expenses they surged by 55% operating income that fell 8% and net income decline by 14%. Earnings per share down 13. And Meta's operating margin collapsed from 43% to 31%. That is a 12 percentage point decline in only one year. Meta produced more than 13 billion of additional quarterly revenue yet generated less operating profit than it did a year earlier. And management while they guided third quarter revenue to between 61 and 64 billion midpoint around 62.5 billion still represents strong growth but it didn't provide the upside investors wanted while Meta's expenses are rising very aggressively. They also raised the bottom of its 2026 capital expenditure range. Management now expects to spend between 130 billion and 145 billion compared with its previous range of 125 to 145. The lowest possible spending outcome has therefore become 5 billion more expensive and that explains the initial reaction. Investors received a modest revenue beat, a substantial earnings miss, collapsing margins, guidance that failed to clear the bar, and confirmation that Meta's enormous investment cycle is not slowing. But this is where the story becomes much more complicated. Meta's total revenue, well, that increased from 47.5 billion from the same quarter last year to now sitting at 60.8 billion. Growth was broad-based across every major geography. US and Canada that reached 27 billion. Europe around 14. Asia-Pacific reaching 11.2. And advertising revenue alone reached 59.4 billion. That's up from 46.6 billion just 1 year ago. That's around 27% growth from an advertising business that was already operating at extraordinary scale. This is not the financial profile of a platform that's losing relevance. And family daily active people that increased around 3.6 6 billion versus 3.48 billion a year ago. Growth of around 3%. It may not sound spectacular, but Meta already reaches close to half the world's population every single day. Explosive user growth is no longer necessary. But honestly, what matters much more is monetization. Average revenue per person increased around $16.90 from 1365 in Q225. That growth is around 24%. Meta is not merely just adding users. They're extracting substantially more revenue from every person already using the platform. And worldwide ad impressions, well, they increase 14%. Impressions, they rose 9% in the US, 13% in Europe. We can see 17% in Asia Pacific and 12% in the rest of the world. So, engagement and available advertising inventory that remains healthy everywhere. And then look at this. At the exact same time, the worldwide average price per advertisement that increased by around 12%. Prices were up 20% in the United States and Canada. So Meta delivered more advertisements and charged substantially more for many of them. That is an exceptionally powerful combination. And Family of Apps revenue reached 60.4 billion this quarter. While Realy Labs that generated only 431 million almost the entire company therefore continues to be supported by Facebook, by Instagram, WhatsApp and Messenger and the core platform remains extremely successful. I mean Meta's quarterly revenues increased from around 21 billion at the end of 2019 to more than 60 billion today. The latest result was the company's highest non-h holiday quarterly revenue in its entire history. The demand problem investors fear, well, that's not appeared yet. Since 2016, quarterly revenue has increased from around 6 billion to more than 60 billion. That is a 10-fold increase. The advertising business has survived competition, privacy changes, recessions, regulation, and numerous supposed threats to Facebook itself. And AI, well, is already contributing to that advertising engine. We can see at the top here, in fact, management have said that 9 million small businesses are now using at least one of Meta's AI powered advertising creative tools. Those tools help businesses create, test, and optimize advertising content much more efficiently, and they produce 59.4 billion of advertising revenue, 49.5 billion of gross profit, and 15.8 billion of net income. The problem is not generating revenue. The problem is what happens to that revenue after Meta earns it. We can see that research and development expense increased from 27% to 36% of revenue. General and admin costs well they rose from 6% to now sitting at 9 and cost of revenue moved from 18% to 19. Almost every major expense category grew faster than revenue and research and development alone reached approximately 21.7 billion during this quarter. Royalty Labs will as we can see here produced a further operating loss of around 4 billion and Meta generated nearly 50 billion of gross profit but only 18.8 of that survived as operating profit. Now net income it fell from 18.3 billion in Q225 to now sitting at 15.8. It remains an enormous amount of profit but the direction matters. Revenue grew 28% while net income declined 14%. That's not operating leverage. it significant operating deleveraguted earnings per share while that fell from $714 to now sitting at $618. And because the market expected more than $720, the miss was considerably worse than the year-over-year decline alone suggests. But even this was not the number that worried me the most. Their quarterly cabex surged from 17 billion to now sitting at more than 31 billion. That's an increase around 83% in one year. During the first half of 26, Meta spent almost 51 billion compared with approximately 31 billion during the first half of 2025. And the longerterm chart shows how extraordinary that acceleration has become. For years, Meta's quarterly investment remained below 10 billion. It then rose to 14 then 1721 and now around 31 billion in a single quarter. Now, Meta still generated around 31.9 billion of operating cash flow, but purchase of property and equipment that consumed just shy of 31 billion, while finance lease payments consumed another 962 million that left only 784 million of free cash flow. And one year ago, Meta produced 8.5 billion of quarterly free cash flow. During the previous quarter, it generated 12.4 billion free cash flow. therefore fell around 91% year-over-year despite revenue increasing by 28%. This is the real earnings warning. The key question is whether investors should ignore this temporary collapse or treat it as evidence that Meta's returns are deteriorating. And this analyst is going to explain why the answer depends entirely on the return Meta eventually produces from this investment. >> Um I just want to ask you about cash flow. I mean the expectation it was going to come in negative for the quarter. It came in at 784 million with an M, which is uh incredibly low for compared to what they had normally been posting in quarters past. Are you concerned at all about that or is that just something you can look through? I I mean I I think the the the free cash flow story for Meta and for all the other hyperscalers goes handinhand with the ROIC story and investors will be willing to accept the the pullback and free cash flow if we're getting negative. Uh obviously to only to certain point um but but it's it's the free cash pullback, the capback spend uh debt and equity raises. uh investors will be willing to be patient if they see the story on the ROIC side coming through as well >> and that is the problem. Meta has shown investors exactly how much money it intends to spend this year but is not shown when the financial return will become visible and management is not even providing a specific 2027 capital expenditure outlook. Now, we know Meta expects to spend between 130 and 145 this year. But we don't know whether 27 represents the peak, another increase or merely the beginning of a much longer infrastructure cycle. That uncertainty deserves a lower valuation multiple. And to understand the investment, we need to understand Zuckerberg's motivation. Meta built one of the largest companies in history, but it doesn't control the devices through which people access Facebook, Instagram, or WhatsApp. Apple and Google control those platforms. >> Actually, I was in the room when he he had a sit down interview with our very own Emily Chang a few summers ago and he actually said, you know, I would much rather overspend and and be wrong betting on this than uh wake up in a few years and realize we underinvested. And so he's willing to lose a little money uh to make sure they're in position. Now you have to remember what's driving this a little bit for Mark Zuckerberg is that he's someone who has built uh his entire business on the backs of other platforms for distribution. So what I mean by that is you know you can't access Facebook and Instagram unless you got an iPhone or an Android phone. That makes the strategy understandable. Meta does not want to depend on Apple, Google, OpenAI or Anthropic to reach its users. But a strategically important investment does not automatically become a financially attractive investment. Shareholders still need an adequate return. And as we said, there are early indications of monetization. 9 million small businesses are already using Meta's AI advertising tools. Meta's also launched Meta 1, a subscription offering that provides additional tools and AI features across its applications. and management says its new models can analyze images, improve their own output, and create better advertising variations. If those tools improve conversion rates, advertisers may spend more, allowing Meta to capture the benefit through higher prices and stronger returns. And Meta is also saying it's receiving offers for computing capacity at a significant premium to what it paid. The company is building APIs, business agents, coding tools, and productivity products that creates the possibility of a new enterprise and cloud style revenue stream. And management are also saying here that daily interactions with Meta AI increased by 60% after the product was rebuilt around its latest models. That is encouraging adoption, but engagement is not the same as revenue, and rapid usage does not guarantee attractive economics. And then you've also got the glasses. Zuckerberg believes AI powered glasses could become a new computing surface that users carry throughout the day. Owning that hardware could reduce Meta's reliance on Apple and Google while creating an entirely new distribution platform. But Realy Labs, well, they lost another $4.6 billion during this quarter. The division still generates only a few hundred million of revenue as we can see here. Now investors therefore I would say have every right to question how much capital should be committed before the economics become clearer. And obviously the optimistic interpretation here is that Meta is building several enormous businesses simultaneously. The pessimistic interpretation is that it's highly profitable advertising platform is funding another collection of expensive projects with uncertain commercial demand. For now both interpretations remain possible. The first major risk well that's duration. If spending peaks during 2026 or 2027 and then normalizes, today's free cash flow collapse may be temporary. If spending continues accelerating through 2028, current valuation models, well, they could be far too optimistic. And you've got the second risk, which is the markets become considerably more selective about AI spending. Several semiconductor companies suffered enormous declines while Microsoft and Google rose after demonstrating clearer revenue growth from their AI and cloud investment. And Microsoft's reaction here demonstrates what investors currently reward. The market will tolerate huge capital expenditure while the revenue and profit attached to that investment are visible. Meta's revenue while it's growing but investors cannot isolate the incremental financial return generated by AI. So that raises an important question. Why did investors reward Microsoft and Google while punishing Meta even though all three are spending extraordinary amounts on AI infrastructure? The difference is not simply the size of the spending is the visibility of the return. Google they can point directly to cloud revenue, infrastructure demand and sale of its custom chips. Microsoft can point to Azour enterprise AI subscriptions and C-pilot. Meta still deres almost all of its revenue from advertising. This analyst explains why that distinction matters. >> Somewhat similar to what we saw with Alphabet. They announced that they were going to be spending more. Shareholders didn't like that too much. Is Is this just part of that trade? >> I I think that's exactly what it is. Investors trying to digest uh what's acceptable and whose position where. I I thought the the alphabet print was was actually uh stronger, right? and and you saw the pullback there. Uh maybe the commentary on the capex spend was was was a little bit more. Google raised the high end of the capex uh spend expectations too versus Meta just on the low end. Uh but but Google had really strong uh cloud business. They started selling their their TPUs. And so they have some of those factors outside of the core search business that that are really pushing right now and accelerating. Um I I think I think for for Meta those factors aren't there yet and investors need to see more uh from that. >> Now that doesn't mean Meta has no additional opportunities. Management's discussing compute sales APIs, business agents, subscriptions and productivity tools, but those businesses are still promises rather than meaningful financial segments. Until that changes, Meta spending will be treated as less proven. and Wall Street, while they reacted immediately. Barkclays lowered its price target to $780, citing concerns about the timing of Ma's returns. Bank of America, well, they also reduced their target to $810 because of higher expenses. Goldman Sachs lowered it to $725, focusing on infrastructure spending. And we have Wolf Research, they reduced their target to $700 because of elevated capital expenditure concerns. And what's interesting to note is that even with all of these cuts, there is still quite a lot of upside from where the stock sits today. The range from analysts is 580 on the lower end, 800 around that on average, with the higher, more bullish analyst coming in just over $1,000. And the bullish case is that Meta remains capable of generating more than $30 per share of annual earnings, while the collapse has pushed its valuation towards levels normally associated with much lower businesses. Here's the simplest version of that argument. >> Yeah. I mean, remember this this company's going to do 30 plus dollars of earnings power. So, you don't have to put a big multiple on the stock to get a higher price target, right? If you you just put a 20 multiple on it, you're over $600. You know, we think it could sustain a higher multiple and it could actually do more than 30. It can be $35 of earnings power. So, you can get to, you know, $8 $900 pretty quick on on a pretty reasonable multiple. The reason why we're excited about Meta is it's the most hated large cap internet name. We think there's more gas in the tank. We think there are more things they can do in in the initiatives that they're doing that can generate these types of returns. That's why the market's current assumptions matter. We can see the estimates from analysts for 2026 as well as the full year of 2027. And if we use the current share price today, well actually these numbers should be updated and based on 2027. Well, Meta is trading below 15 times 27 earnings. And in fact, when we look from Simply Safe Dividends before the full post earnings decline, Meta looked to be trading around 18 times forward earnings. Well, at around the share price drop today, it's sitting much lower around the 16 region. that is significantly lower than their 5year average around 22. And even based on the unupdated valuation drop, this was showing signs of undervaluation where it was below the blue tunnel which here indicates the intrinsic fair value price. And this is an important qualification. Meta still trades above the median company in its sector on several metrics which explains why it does get a weak sector relative valuation grade. But comparing Meta with a median communication service company is not really meaningful if you want to look in terms of including growth. While Meta now trades on a PG below one at 85 32% cheaper than the sector and 44% below their own 5-year which typically sits at 1 and a half. Meta also as we can see from seeing Alpha gets an A+ profitability grade. Its trailing gross margin sitting at 82%. Its net income margin sitting at 33% and return on equity sitting around the 33% level. Very few companies of any size produce this level of profitability and revenue growth. Well, it also remains dramatically above the sector. The major weakness here is in fact the free cash flow as we can see down 32% year-over-year. So the valuation debate is not whether Meta is a highquality business. It's whether the investment cycle permanently reduces the cash available to shareholders today. And over the last 5 years, Meta produced a total return of around 65% pretty much identical to the S&P 500 but below the NASDAQ. The period includes the 2022 collapse, the year efficiency, and the current AI investment cycle. Volatility around Zuckerberg's spending is not a new thing. And over the last 10 years, while Med has returned around 372%, outperforming the S&P 500 but trailing the NASDAQ, the company's repeatedly rewarded investors who bought when spending fears overwhelm confidence in its core business. But historical recoveries do not guarantee this one. Now, let's move to my updated discounted cash flow model. I want to be transparent because valuing Meta using current free cash flow alone would be almost meaningless. The company generated only 784 million this quarter because investment spending is unusually elevated. This model therefore assumes extremely weak cash generation during the peak investment period. I've used roughly in line with analysts only 2 billion for 2026 and negative 20 billion for 2027. The assumptions deliberately recognize the possibility that spending becomes even more aggressive. And the negative 2027 assumption is not a prediction that Meta will definitely burn 20 billion. It's a stress assumption reflecting the fact that management has not told us where spending settles next year. This is the largest uncertainty inside the valuation and the model assumes 2026 and 2027 represents the most intensive infrastructure years. From 2028, capital expenditure begins normalizing while the assets Meta has built start producing meaningful economic returns. Without that normalization, the valuation would need to be substantially lower. And then I've used 80 billion of free cash flow in 2028 followed by a 12% annual growth in the central scenario is a sharp recovery and is the most aggressive part of the model. It requires Meta's infrastructure cycle to genuinely peak. However, Meta already generated almost 32 billion of operating cash flow during this quarter alone. The business has not lost its ability to produce cash. management is choosing to reinvest almost all of it. The valuation depends on how much of that spending becomes permanent. Now looking at the three different scenarios under the low growth using 10% long-term growth, the model produces an intrinsic value of $716, the central middle one at 12% 788 and the more optimistic at 14% $866. So the central intrinsic value estimate coming in at $788 compared with the market price today is indicating around a 32% margin of safety and Wall Street well based on their target of 798 not too dissimilar from where we put our target that implies around 49% upside and the reverse CCF is equally interesting at around $536 the market appears to be pricing in only 4% long-term cash flow growth under these assumptions. For a company currently growing revenue by 28% that initially appears undemanding with the market not assuming advertising revenue suddenly stops growing is assuming that enormous capital expenditure absorbs a much larger percentage of Meta's operating cash flow permanently. That's a central disagreement between bulls and bears. And if Meta's AI tools improve advertising performance, accelerate product creation and generate subscription revenue, the investment could strengthen its existing economic mode. In that scenario, even the $788 central valuation might eventually prove conservative. And if Meta successfully sells computing capacity, APIs, agents, and productivity products, it could also develop a major revenue engine outside advertising. That is precisely what investors currently believe is missing compared with Microsoft and Google. But if spending continues rising and those products fail to produce meaningful revenue, free cash flow may remain suppressed for much longer. In that case, the 2028 normalization inside my model would be too optimistic. The $788 estimate, well, it's not guaranteed. So, three conditions must be met for the central valuation to become realistic. First, the investment cycle eventually needs to peak. Medic cannot permanently spend almost all of its operating cash flow on infrastructure while retaining the valuation of an asset like technology platform. Second, operating cash flow must continue growing while free cash flow recovers. The bullish thesis is not that Meta suddenly stops investing. It's that revenue and operating cash flow eventually while grows faster than capex allowing meaningful cash to reach shareholders again. And third, meta needs to demonstrate that AI creates incremental economics. That can mean better advertising conversion, higher prices, subscriptions, enterprise products, or compute sales. The model does not require everything to succeed, but it requires more than a more expensive version of today's business. And honestly, the first warning would be another major increase in 2027's capital expenditure expectations without clear evidence of incremental revenue. At some point, uncertainty stops being temporary and becomes part of the permanent economics of the company. The second warning would be free cash flow remaining close to current levels for several consecutive quarters. One unusually weak quarter can reflect the timing of infrastructure purchase. Repeated weakness would suggest Meta has structurally become a much more capital inensive business. And the third warning would be deterioration inside advertising itself. If impressions slow, pricing weakens or revenue per user stops rising while expenditure remains elevated, the investment case changes completely. Right now, the core business is compensating for the spending cycle. And at around $536 where MetaTrades not too far off its 52- week low, the stock sits below its historical earnings multiple below revised Wall Street targets and substantially below my central DCF estimate. That makes the riskreward attractive. But the underperformance has also become extreme. Before this latest decline, Met had already fallen around 16% over 1 year while the S&P 500 gained around 15%, the Nasdaq more than 17%, the valuation reset while the core business continue growing. Now, it doesn't mean the stock cannot fall further. The market may remain focused on spending until management provides credible 2027 guidance or free cash flow begins recovering. An inexpensive stock can remain inexpensive while uncertainty is increasing. But we should separate the company from the project portfolio. The family of apps remains one of the finest businesses in the world. It generated 23.4 billion of quarterly operating profit. Despite Meta's enormous investment in AI and realy labs and users are growing, revenue per person increased by around 24%. Ad impressions rose 14. Average advertising price increased 12. The underlining economic engine is performing exceptionally well even while reported earnings and free cash flow decline. However, the reason I wouldn't describe Meta as a risk-free bargain is that shareholders do not yet know how much of this spending is temporary at 130 to 145 billion of annual capex. Small forecasting errors can remove tens of billions from free cash flow. So, my conclusion is that Meta has entered a genuine buying range around $536. I classify the stock as a buy, not an all-in buy. The valuation compensates investors for significant uncertainty, but it doesn't eliminate the uncertainty. I'd begin building or adding to a position around these levels rather than assuming I can identify the exact bottom. The low scenario 716, the middle 788, and the high at $866 with Wall Street's average price target $798. Remarkably close to my updated valuation. Now analyst targets are not intrinsic value and they frequently lag changing conditions but the agreement reinforces how pessimistic the current market price has become. The single most important number from here is not daily active users or even quarterly revenue. It's the 2027 capital expenditure outlook that will reveal whether the current free cash flow collapse is temporary or becoming structural. So Meta's stock has fallen because the market can no longer ignore the cost of Zuckerberg's AI ambition. The earnings missed, weaker margins, elevated capex, and only 784 million of quarterly free cash flow all justify greater caution. The decline here is not irrational, but the core business is not collapsing. Advertising revenue grew 27%. Meta reaches 3.6 billion people each day. It's delivering more advertisements, charging higher prices, and producing tens of billions of dollars in operating cash flow. and around $536. I believe the markets move from appropriately cautious to overly pessimistic. My central estimate $788, but reaching that value depends on CABIX eventually normalizing and AI investment beginning to produce measurable returns. So, my answer is straightforward. Meta is now a buy for long-term investors who can tolerate spending uncertainty and further volatility. I'd scale into the position rather than purchase everything at once. And I'd monitor capex more closely than any other metric. Ultimately, the market believes Zuckerberg may be overspending on another expensive vision. Zuckerberg believes he's securing Meda's place in the next computing platform. At a 52- week low and around 16 17 times expected earnings, investors are now finally being paid to take that risk. But let me know your own thoughts in the comments below. Don't forget to sign up to the weekly newsletter where we drop one once a week covering severely undervalued stocks and what's going on in the market. More importantly, have a great day.