Disclaimer: This is not financial or investment advice. I'm sharing my personal investment decisions and reasoning for educational and informational purposes. Always do your own research before making any investment decisions.

We've all seen the headlines.

Microsoft is in one of its largest drawdowns of the last decade.

The stock is down 32%, its second worst drawdown of the past ten years, and just a hair behind the 2022 tech slowdown that saw Microsoft fall 36% at its peak. With this recent decline, Microsoft now trades at one of its lowest valuations over the past 10 years, at just 21x forward earnings.

Many investors claim this is a clear buying opportunity. You get the chance to own one of the leading Magnificent 7 at a "cheap" valuation, or so it seems. Just about every analyst on Wall Street is bullish today.

A real bargain shouldn't need a calculator. If the price is right, the margin of safety does all the work for you.

Warren Buffett said it best:

The opposite approach to what Warren Buffett described is to model your way to certainty. It rarely works.

Many analysts try do this. They pick a business, break out every revenue segment, calculate growth rates for each, attach expected margins, and layer on a myriad of other factors. Essentially, they model every line of the three financial statements to try to perfectly calculate the intrinsic value.

So which is Microsoft? At 21x forward earnings, is it the kind of setup that screams cheap, or is it a trap that will lead to flat returns for a decade? The last thing we want is something that looks like this:

The purpose of this exercise is to run the rough math. Not to pin intrinsic value to the dollar, but to get a sense of what Microsoft is worth.

The exercise needs two inputs. What the business can earn in the future, and what the market will pay for those earnings. We begin with the first.

What has Microsoft's diluted earnings per share grown at?

As we can see, Microsoft has comfortably compounded earnings in the 20% range over the years.

At the end of Fiscal 2025, Microsoft's diluted EPS was $13.64. I'll use this as the baseline. Analysts expect EPS to compound at roughly 18% from Fiscal 2025 to Fiscal 2028.

Using 18% as the anchor for years 1-5, I'll pull the assumption down slightly for the base case to stay on the safe side. For years 6-10, I'll account for modest deceleration, since it gets harder to grow at scale.

Now, onto what the market will pay for Microsoft’s future earnings.

Historically, Microsoft has traded at a median forward P/E of 28x over the last decade, with a low of 17.5x and a high of 37x. Given that the market now recognizes Microsoft’s durability and competitive advantage, it's unlikely we’ll revisit that 17.5x low. That said, companies do go through nightmare scenarios. Remember, Meta dipped to 8x earnings near the end of 2022. Yes, 8x earnings. Sometimes it's crazy to think about, but it's true.

Assuming no nightmare scenario for Microsoft, I'll use 18x as the terminal multiple for the bear case, 21x for the base case, and 24x for the bull case. This assumes no major multiple expansion, which adds another layer of conservatism.

Putting it all together, and to lock in a 10% annualized return over the next decade, I’d need to buy Microsoft at $629.

The stock trades at $371 as of April 10th, 2026. If these growth assumptions and terminal multiples hold, Microsoft is priced to deliver ~17% annualized over the next decade. Conservative inputs, above-market returns. Time to pick out the color of the Ferrari F80, right?

If only investing were that easy.

The reality is, Microsoft is evolving into a different business. Not because its segments are changing, but because of the capital intensity brought by the AI buildout. Most investors and analysts are focused on what Microsoft’s stock will do in the next year or two. They aren’t thinking about what happens when nearly $1 trillion of capital expenditures hits the income statement over the next decade.

And that changes everything. Here’s why: