Morgan Stanley: Alphabet's AI Capex Could Reach $375 Billion by 2028
Taylor Wilson
Morgan Stanley raised its Alphabet capex forecast for 2027 by 17% to $350 billion and sees 2028 reaching $375 billion — roughly double this year's expected spend — exposing a core tension where AI investment is outrunning operating cash flow.
How big are these numbers, really?
Analyst Brian Nowak lifted Alphabet's 2027 capex estimate by 17% to $350 billion, with 2028 projected at $375 billion.
This means → annual capital spending two years out will be roughly double this year's level, an expansion pace far beyond normal corporate investment.
Alphabet had already signaled a "significant increase" over 2026's up-to-$190 billion ceiling. Morgan Stanley's estimate puts a hard number on "significant."
Where does the money come from? Can cash flow cover it?
Alphabet completed an $85 billion equity raise in early June — a clear sign that operating cash flow alone cannot fund the build-out.
In plain terms = the company's own earnings are no longer enough, so it has to tap the market — either borrow or issue new shares.
This reflects a structural turning point for Big Tech's AI expansion: capex growth now outpaces cash-flow growth, deepening reliance on external financing.
What does all that spending buy?
Morgan Stanley estimates the investment will add roughly 9 gigawatts of compute capacity by 2028 — a gigawatt measures data-center power scale, enough to supply about 700,000 homes.
On the revenue side, the bank forecasts Alphabet's total revenue growing just 4% in 2027 and 1% in 2028 — modest at best.
Google Cloud, however, is the outlier: projected to grow 106% in 2027 and 44% in 2028, making it the primary revenue engine. This means → Alphabet is concentrating its bet on cloud.
Why is the stock falling despite all this?
Alphabet shares have dropped about 7% in June after hitting an all-time high above $400 in early May, erasing roughly $315 billion in market cap.
Nowak's explanation: "investor capital continues to flow toward hardware and semiconductor names" — the market currently prefers to own the pick-and-shovel sellers, not the miners.
In plain terms = investors trust the tangible hardware links of the AI supply chain more right now, and are staying on the sidelines for platform companies like Alphabet that spend big and wait for returns.
What is Morgan Stanley's bottom line?
Target price raised from $375 to $415, with an overweight rating maintained — implying roughly 16% upside from Tuesday's ~$356 close.
Nowak argues that "fundamentals and visibility for 2027 and 2028 are improving," making the pullback a tactical buying window.
The core tension, however, remains unresolved: skepticism over near-term returns versus expectations for long-term cloud growth will be tested across earnings reports over the coming quarters.
Content is for reference only, not financial advice.