Most investors rarely think about how many shares a company has outstanding. You buy a stock, and your ownership stake stays the same.
That changes when a company issues new shares. The business now has more shares outstanding, meaning each existing share represents a smaller ownership stake.
Meta Platforms (META) averaged 3.64 billion daily users across its apps in September, while third-quarter revenue climbed 26% to $64.6 billion, according to its earnings release.
Read more: Meta wants users to pay $249 to put AI on their face
“We hit major milestones this quarter toward delivering personal superintelligence for everyone,” said CEO Mark Zuckerberg in the release.
However, a subtle shift has emerged in how Meta pays for that ambition.
According to a Bank of America research report seen by TheStreet, Meta’s plan to sell up to about $100 billion of stock could trim its earnings per share by about 6%.
Meta’s stock sale could cost shareholders about 6%
Meta announced the plan on Oct. 26. It priced $25 billion of Class A stock at $625 a share, upsized from $22.5 billion, plus $12.5 billion of 6.5% mandatory convertible preferred stock, according to its pricing announcement.
It also plans a separate program to sell up to $60 billion of new nonvoting Class C shares, starting no earlier than the first quarter of 2027.
Here is how the pieces fit together, according to Meta’s filings:
Class A common stock: $25 billion at $625 a share, sold through forward agreements that settle by Nov. 19, 2027
Convertible preferred: $12.5 billion, paying 6.5% and converting into common stock around Aug. 15, 2029
Underwriter options: Up to about $5.6 billion more across both deals
Class C program: Up to $60 billion, with no agreement signed yet
Nvidia (NVDA), AMD (AMD) and Broadcom (AVGO) indicated interest in buying up to $5 billion of the Class A shares, though “these indications of interest are not binding,” Meta said.
Analyst Justin Post laid out the cost in a note titled “Anticipation becomes a reality, Meta announces equity raise.”
Assuming about $95 billion is raised near $640 a share, “we estimate Meta would issue roughly 150mn shares, which would be ~6% dilutive to EPS,” wrote Post. That is up to about $2 off his 2027 earnings estimate of $34 a share.
By my math, 150 million new shares on top of roughly 2.57 billion outstanding means about 5.8% more shares splitting the same profits.
The stock closed at $630.42 on Oct. 27, according to price history from InvestSmart. That’s well below its 52-week high of $779.82, the note showed.
Meta plans to raise up to about $100 billion through stock and preferred sales to fund AI.NurPhoto / Getty Images
Why the AI buildout needs outside money
The reason is simple. Meta’s AI spending now outruns the cash its ads bring in.
Meta expects $135 billion to $145 billion in capital spending this year, the release showed. Free cash flow swung to negative $2.12 billion in the third quarter from positive $10.63 billion a year earlier, TheStreet reported.
The buildout is enormous. Meta “is building four superclusters with more than one gigawatt of capacity each” and is on track to “nearly double total infrastructure capacity in 2026,” the company said.
More on Meta and AI:
Bank of America revamps Meta stock price target once again
Zuckerberg says Meta’s AI bet only fails if all AI fails
Louis Navellier sends strong message on Nvidia’s circular deals
Post expects the bill to keep growing. He estimates $210 billion in capital spending in 2027 and $32 billion in negative free cash flow, and he wrote that some investors are “around $240bn in capex, and others are even higher.”
JPMorgan has projected $243 billion of Meta capital spending for 2027, TheStreet reported.
Given that backdrop, it’s easy to understand why Post thinks the raise was coming.
“The financing potentially being raised at $95bn suggests higher capex than BofA/Street estimates in 2027,” he wrote.
Bank of America says AI revenue has to catch up
The question now is whether the spending pays off.
Post flagged three main investor worries: higher and longer-lasting capital spending, “dilution overhang” from the $60 billion Class C program, and “the risk that AI infrastructure investment scales faster than monetization.”
He named the proof points to watch. Those include adoption of Meta’s Muse AI agent, Business Agent monetization and enterprise AI [selling AI tools to companies rather than consumers].
Muse drew more than 10 million U.S. and Canadian sign-ups in its first six weeks, TheStreet reported. Ads still pay the bills, though, and Post expects Meta “to push the ad business to support the stock price.”
Much will likely depend, however, on whether those new products turn into revenue before 2027’s spending lands. A year ago, Meta was generating more than $10 billion of free cash flow a quarter.
I’d also note an awkward detail. Nvidia, AMD, and Broadcom sell Meta the chips this money will help buy, and they’ve now signaled interest in buying Meta stock.
What dilution means for your Meta shares
If you own Meta directly, or through an S&P 500 index fund, your share of future profits will shrink a little as new stock is issued.
That isn’t automatically bad. Dilution pays off if the new money earns more than it costs, and Post’s math suggests stock is a slightly cheaper option than borrowing.
A $100 billion debt raise at 7% interest “could be 6.8% dilutive to 2027 EPS,” while a similar equity raise “would only be 6.5% dilutive,” Post wrote.
For Post, the trade-off still favors owning the stock. He kept an $845 price objective, about 32% above the $638.06 price in his note.
“Given a big AI opportunity ahead, we maintain our Buy rating,” he wrote.
More on Meta & its stock:
Is Meta Platforms a good long-term investment?
Does Meta pay dividends? Its yield and payouts explained
Has Meta conducted a stock split? What sets this ‘Mag 7’ stock apart
Meta’s stock buybacks: How the company’s AI spending could affect shareholder returns