The article is an examination of how to study companies’ financial disclosures so that you can understand how all the pieces fit together. Here at Calcbench, however, we just want to remind everyone — we have all that data there for the taking, with just a few easy clicks.

Let’s start with a recap of the article itself, written by columnist Jonathan Weil. The premise is that while Meta is generating heaps of operating cash flow, which would seemingly help to cover the huge capital expenditures Meta is laying out to build AI data centers, that’s not how it works in practice. 

In practice, because Meta must cover the cash costs of equity compensation given to employees, and those costs are expensive, the company doesn’t have heaps of cash that can cover the costs of data centers. That’s why Meta is taking on new debt to pay for its data centers, even as operating cash flows increase.

This phenomenon — of employee-related equity compensation costs gobbling up free cash flow — exists at lots of tech companies, as Weil’s column notes; Meta just happens to be an outlier example.

With Calcbench, however, analysts can perform your own analysis of this issue quickly and easily. Then you can bring better judgment to your own assessments of whether and how the AI hyperscalers can afford all this.

Let’s walk through a comparison of Meta ($META) and Alphabet ($GOOG). 

Finding the Relevant Disclosures

Finding the relevant disclosures in Calcbench is straightforward. Just go to our Multi-Company page, call up the companies you want to research (in this case, Meta and Alphabet), and enter the relevant disclosures in the Search Standardized Metrics field on the left side of your screen.

For our purposes today, the six metrics we want to study are:

• Operating cash flow

• Free cash flow

• Gross capital expenditures

• Value of shares repurchased during the period (not the number of shares repurchased, which Calcbench also tracks for you)

• Payments related to taxes for share-based compensation

• Total debt

We entered all those metrics, and Calcbench immediately returned Figure 1, below. You should see something similar on your screen.

Among those six metrics, the only one that’s a bit tricky is the payments related to taxes for share-based compensation. Both Meta and Alphabet do list that number in the earnings release, specifically on the Statement of Cash Flows under financing activities — but they tag the disclosure somewhat differently, so you might need to check the actual earnings release to confirm. But the number will always be there somewhere. 

Anyway, once we pulled together the data and lined them up in a spreadsheet, we arrived at Figure 2, below.

Back to Weil’s column in the Wall Street Journal. His key point was that when you add Meta’s repurchase spending and equity withholding taxes together, they equal nearly 97 percent of free cash flow. So Meta actually doesn’t have oodles of spare cash lying around to build data centers; it needs to preserve the $115 billion in operating cash flow for operations, rather than just capex. That’s why the company is tapping debt instruments to pay for its data center dreams. 

By comparison, Alphabet’s repurchase spending and equity withholding taxes are only 81 percent of free cash flow. One could run the same comparison exercise for Amazon ($AMZN), Microsoft ($MSFT), and Oracle ($ORCL) too, although beware that Microsoft and Oracle run on July 1 and June 1 fiscal years, respectively, so their annual report data is rather out of date by now.

So the issue for Meta is that its spending on equity compensation is consuming operating cash flow, which in turn narrows the company’s options to finance its AI data center dreams and drives it toward more debt. Little surprise, then, the Financial Times just had an article too, noting that Meta is cutting staff equity awards for the second year in a row. 

If you’ve been using Calcbench for financial analysis all along, you could’ve seen that coming!