How much does the AI buildout actually have to pay off? I solved each hyperscaler's price backwards

Over the last couple of weeks, I posted DCFs on the three big hyperscalers - [Microsoft](https://www.reddit.com/r/ValueInvesting/comments/1uio0by/six_months_ago_we_called_microsoft_overvalued_at/), [Meta](https://www.reddit.com/r/ValueInvesting/comments/1umbmng/the_meta_version_of_my_microsoft_capex_breakdown/), and [Alphabet](https://www.reddit.com/r/ValueInvesting/comments/1upov82/google_looks_cheap_at_27x_earnings_but_a_dcf/). The most common pushback, especially on the Google post, was that I was counting the cost of the AI capex without giving credit for the return it will eventually earn. It was a fair criticism, so instead of arguing about it, I ran the model backwards to actually estimate the numbers. Read the other posts to fully understand the reasoning.

One thing to clear up first: my DCF is not blind to the payoff. It drives revenue off analyst estimates, and analysts have already baked their view of the AI return into those numbers. So the return is in there, second hand, through consensus growth.

So I asked this question: at today's price, how much does each company actually have to deliver to be exactly fairly valued? I solved it two ways, one lever at a time. What if revenue grows faster or slower, and what if the margin changes? Here are teh results (numbers might be slightly different than in my other posts as they are from today's data):

|Company|My DCF value|Price|Revenue growth to be fairly valued|vs 17% consensus|Or EBITDA margin|vs today|
|:-|:-|:-|:-|:-|:-|:-|
|GOOGL|$252|$362|25.7%|\+8 pts|52%|\+14 pts|
|MSFT|$483|$383|12.6%|\-5 pts|46%|\-9 pts|
|META|$637|$603|6.6%|\-1 pt|46%|\-2 pts|

So what does the analysis tell us:

\- GOOGL has to beat the Street. To justify $362, revenue has to compound about 25.7% a year, roughly eight points above the 17% consensus used in the model. That is revenue reaching about $1.27 trillion by 2030, versus the $900B analysts already model. Or the margin has to jump from about 38% to about 52%.

\- MSFT is the opposite. It is already undervalued at consensus, so it has room to spare. Revenue can grow about five points slower than consensus, or the margin can fall about nine points, and it is still fairly valued at today's price. That is the cushion the selloff has built in.

\- META sits almost on the line. A small cushion, about one point of growth or two of margin, and it is fairly valued. The price is close to what revenue growth analysts already assume.

Note that all three analyses are based on the default capex model, which assumes capex tapers off over time.

Disclaimer: This is for educational purposes only and is not investment advice. The author and Stockoscope may hold positions in the securities mentioned. Always do your own research.