"Always acknowledge a fault. This will throw those in authority off their guard and give you an opportunity to commit more."-Mark Twain.

Barron's writes a lot about stocks. Not every article written is our opinion. Oftentimes, we relay what investors and analysts are saying, so readers can integrate updated opinions into their analysis. We pick stocks, too. Offering our views on what will happen to a given stock over the coming 12 months.

With stocks, or anything else, it's impossible to be right all of the time. Our views on SpaceX.and its astronomical valuation certainly weren't.

Shortly before SpaceX's record-setting IPO, we called SpaceX stock worth $90 a share. That was based on several factors, including third-party valuations, one of which was from NYU professor and valuation maven Aswath Damodaran. Before that, we said that SpaceX was the big loser in its February merger with xAI.

At the time of the merger, Elon Musk's AI company was burning cash in an uncertain bid to catch AI leaders such as Anthropic. And SpaceX was "a cash machine," said Greg Martin, managing director at the private market exchange Rainmaker Securities, with industry-leading profit margins and Starlink, an impressive, fast-growing space-based broadband business.

Now, it looks as if SpaceX shareholders were the winners in the xAI deal and that while SpaceX is still hard to value, it's worth far more than $90 a share.

What changed is the AI business. Revenue has simply accelerated faster than anyone expected. That growth has radically altered near-term earnings forecasts and long-term cash flow projections.

Around the time of the IPO, Wall Street forecast 2027 sales and earnings before interest, taxes, depreciation, and amortization, or Ebitda, of $70 billion and $28 billion, respectively. Now, those estimates are $100 billion and $59 billion. Faster growth has cut expected cash burn. In July, Wall Street expected 2030 cash use of $24 billion. Now, Wall Street expects positive free cash flow.

AI has done the heavy lifting. SpaceX's AI business is now expected to generate $60 billion in 2027 revenue, up from a July estimate of $38 billion. By 2031, Wall Street projects $530 billion in AI-related revenue. The earliest estimates for 2031 AI revenue averaged closer to $150 billion.

In June, Damodaran valued SpaceX at about $100 a share. His analysis included 2036 AI revenue of $160 billion. That number now looks way too low. At 2036 AI sales of $500 billion-near Wall Street's 2031 revenue call-SpaceX stock would be worth closer to $140. At $1 trillion in 2036 AI revenue, SpaceX stock would be worth closer to $200.

That math rests on many tacit assumptions, including guesses about profit margins and growth in SpaceX's other businesses. Still, it implies that $100 billion in annual AI revenue in 2036 is worth about $10 a share today.

Barron's isn't calling SpaceX stock a buy yet. Exactly how AI revenue will turn out is hard to say. We're just saying it's worth more than $90. What's more, the current valuation isn't all that astronomical anymore. At current prices, SpaceX stock is trading for about 34 times estimated 2027 Ebitda. GE Aerospace and GE Vernova trade for closer to 25 times.

(The S&P 500 trades for about 13 times. Those three companies are expected to grow faster than the overall market.)

One person who isn't ready to change his numbers yet is Professor Damodaran. Most of SpaceX's current AI growth is coming from renting its computing capacity to others, including Google and Anthropic. "That actually takes away from their AI story, since to win in that story, you have to be generating revenues from creating AI agents and collecting subscription or usage revenues," he tells Barron's.

SpaceX needs to generate AI applications on its own, or risks losing revenue when AI computing capacity isn't constrained. "The analogy I would offer would be to a young, manufacturing business that is building a large factory for what it contends will be a huge growth business, but then chooses to lease out three quarters of the factory to their biggest competitors."

It's a fair point, and something for investors to watch. Still, the size of the AI business and opportunity is bigger than almost anyone imagined-Barron's included.

As Mark Twain says, we'll admit that error in forecasting, so we can move on to the next thing we don't get quite right.