Why Alphabet Fell 7% Despite an 82% Cloud Surge

Almost every part of the company fired at once

Alphabet's quarter was one of the cleanest big-tech prints I have read in years. The stock fell 7%.

Start with the results. Revenue jumped 24% to nearly $120 billion, crushing estimates. Growth like that at this scale is genuinely unusual — a four-trillion-dollar company moving at a startup's growth rate.

The detail was strong too:

• Search grew 17%. That is the opposite of the fear that dominated the market a few years ago. Google says its new AI answers are actually increasing searches, not killing them.
• Google Cloud exploded 82%, with profits more than tripling. This was the star of the show.
• YouTube ads and the subscription business kept growing nicely.

And the stock still fell. Two reasons, both worth understanding.

Reason one: $99 billion of that $112 billion profit was not real

Of the jaw-dropping $112 billion headline profit, roughly $99 billion came from a one-time investment gain.

I was covering the release live, and the moment the number printed I thought: this has to be a mark-up on something like a SpaceX stake. Under current rules, if you hold a stock and it appreciates, you report that appreciation in income even if you never sold a share.

So the sentence "Alphabet earned $112 billion this quarter" is wrong. Some of it was operations. Most of it was a bookkeeping entry. It is structurally identical to what Tesla did in the same week with roughly $1 billion of paper gains. I broke down how this distortion works in Why a $112 Billion Profit Can Be an Illusion: How Unrealized Gains Distort Earnings.

Reason two: $45 billion in a single quarter, and the first cash burn ever

This is the bigger one. Alphabet spent $45 billion on AI in a single quarter — more cash than the entire business brought in during that period.

The result: for the first time ever, Alphabet burned cash instead of printing it. This from a company that has been the definition of free cash flow for two decades.

This is where serious investors split. Warren Buffett has been loading up on Google. Bill Ackman just sold his entire stake for a large profit. Same numbers, opposite conclusions.

The bull case, and it is powerful

One: AI is making search better, not worse. AI mode already has over a billion users every month. More importantly, it is finally helping Google monetize the longer, more complicated questions it always struggled to make money on. The thing everyone feared would kill Google may actually be strengthening it.

Two: Cloud is becoming a second search-sized money machine. It is growing over 80% a year, profits are soaring, and it already carries a $500 billion backlog of signed business waiting to be delivered. That is an entire second engine bolted onto the company.

Three: distribution nobody on Earth can match. Google can drop its AI in front of billions of people instantly through Search, Gmail, Android, Chrome, YouTube, and Maps. Nearly 90% of America's largest companies already use its AI tools. A startup would give anything for that kind of built-in reach.

What the numbers actually say

Market cap is $4.05 trillion; enterprise value is $4.09 trillion. The gap is about $40 billion — nothing for a company that generated $53 billion last year, though the heavy capex quarter is worth keeping in mind.

The stock trades at 16.5x earnings and 76x free cash flow. The reason those two multiples diverge so wildly is exactly the two issues above. Net income of $244 billion includes roughly $100 billion of SpaceX gains, which makes the P/E look artificially low. The capex surge compressed free cash flow, which makes the price-to-FCF look artificially high. Both numbers are distorted. Remember that for the next year or so.

What impressed me more was the underlying profitability once you strip the investment gain out. Net margin over the last ten years: 30%, then 34% over five, now 36–37%. It keeps getting better. This was a company in the 20% range not many years ago. Revenue growth is remarkably steady too: 18.5% over ten years, 15% over five, 15.5% over three. That is what happens when one company owns the two largest search engines in the world — google.com and YouTube, which it bought for $1.6 billion.

Return on invested capital runs 17% over seven and five years. The trailing year dipped toward 10% as cash from operations fell, and I do not believe that is permanent.

On the eight-pillar checklist, the only misses are free cash flow and the multiples. Low debt, rising revenue, rising net income, high returns on capital, and shrinking share count all pass.

Whenever multiples come up, I ask the same question. If two companies are identical, but one grows 20% a year and the other grows 5%, which one deserves the higher multiple? The multiple is not the answer. How long the growth lasts is the answer.

Valuation: my assumptions and what came out

Analysts model $14.50 per share this year, more than doubling to $31 within five years — about 15% annual growth — with revenue going from $500 billion to $1 trillion over seven years, roughly 10% a year.

Here are my ten-year assumptions. Revenue growth of 7%, 9%, and 13%. Net margins of 28%, 30%, and 32% — excluding SpaceX and every other equity investment, purely the fundamentals of the business. Exit multiples of 20x, 23x, and 26x. A 9% required return, the same as I used for Tesla, to find intrinsic value.

The margin assumptions may actually be conservative, since the last five years came in at 33%. The multiple is more debatable, and my view is that Alphabet does not deserve a market-average 15 or 16. Search, YouTube, Android — this company is embedded in daily life. If someone told me the right number was 25, I would not argue. Deciding how much premium to grant a business woven into everyone's routine is the art rather than the science of investing.

The output: a low of $250, a midpoint of $350, and a high of $560. The stock is at $326.

So with no margin of safety, this sits just under fair value. Which leaves one question for you: how much margin of safety do you need to feel comfortable? I went through the business segment by segment in Alphabet (GOOGL) at $3.87 Trillion: Search, YouTube, Cloud, and Waymo Dissected.

Risks and counterarguments — the bears have real ammunition

One: AI could eat Google's golden goose. Google has handled it well so far, but it is early to call it clear. If an AI hands people a complete answer, they stop clicking ads and stop visiting websites. Ads are still about two-thirds of Alphabet's revenue. The very thing driving usage could quietly shrink its best business.

Two: the spending explosion. Google may have to pour out hundreds of billions just to keep pace in AI, and that money is not free. AI chips wear out and constantly need replacing. That could crush the free cash flow Alphabet was always famous for — which is precisely what happened this quarter. I laid out my broader view of this capex cycle in Is AI Infrastructure Spending a Repeat of AWS or of the Fiber Bubble?.

Three: antitrust. The government won its big case, and the penalties could slowly chip away at the exclusive deals and the mountains of data that made Google dominant in the first place.

FAQ

Q: If the quarter was that good, why did the stock fall? A: Two reasons. Roughly $99 billion of the $112 billion headline profit was a one-time investment gain rather than operating performance, and $45 billion of AI capex in a single quarter caused Alphabet to burn cash for the first time ever.

Q: Isn't 16.5x earnings cheap? A: That P/E is not trustworthy. The denominator — $244 billion of net income — includes roughly $100 billion of equity investment gains. Strip those out and the real multiple is materially higher. Conversely, the 76x free cash flow figure is overstated because of the capex surge.

Q: What distortion should I watch for next quarter? A: The same illusion running in reverse. SpaceX stock is down considerably, so a large unrealized loss could drag net income down even though nothing about the operating business got worse. Judge it on operating income and segment revenue instead.

Q: How should I read the $500 billion Cloud backlog? A: It is business already contracted but not yet delivered and recognized as revenue. It provides visibility into future sales and supports the case that 82% Cloud growth is not a one-quarter fluke. Just remember that recognition timing and margin vary by contract.