As Amazon.com (AMZN) heads inexorably towards producing losses, some analysts are already switching to using "pro-forma" earnings where they ignore compensation costs paid in stock. For instance, Kevin Kopelman from Cowen & Co says:
"Amazon's Q3 non-GAAP operating profit guidance of $(75)-225 million was below our expectations as the company builds out ahead of the Q4 holiday season"
The non-GAAP operating profit ($(75)-225 million) is derived from taking Amazon.com's operating profit guidance for Q3, $(350)-$(50) million and adding back stock-based compensation of $275 million (a small part of this is amortization of intangibles).
Now, usually some doubt is cast on this attitude because we're usually taking about the issuance of stock options, with a value that's subjective and calculated using models such as Black-Scholes. Those options are usually awarded at the money (that is, their strike is the stock price at the day they're awarded), and might only gain in value if the stock goes higher in the future.
However, with Amazon things are even simpler. Amazon.com awards "units", that is, shares of stock. For instance, it granted 4.3 million shares in Q2 alone. Amazon.com then recognizes the cost of awarding these shares to its employees as they vest.
Amazon.com awarding shares and some analyst trying to ignore their cost is thus absurd for a very easy to explain reason: If Amazon sold 1 share into the market and took the proceeds and gave them to the employee, no analyst would ignore the cost of that share. Yet, if Amazon gives 1 share to the employee and the employee sells it into the market (or keeps it) then someone is actually trying to ignore the cost of a situation that's the same in substance. Magical thinking is needed to somehow believe that the second alternative (giving stock to the employees) is materially different from the first (selling stock and giving proceeds to the employee). Under both alternatives the end result is that Amazon.com issues stock and pays the employee, they are the same!
Additionally, it makes no sense to consider cash flow from issuing shares to pay employees. The reason is simple, if AMZN was selling those shares into the market, the cash from the sales wouldn't be considered operating cash flow or free cash flow, either. So Amazon issuing them and using them to pay employees - which is equivalent to selling them in the market - shouldn't be considered either.
While ignoring the cost of options is already a very doubtful practice, ignoring the cost of shares that are given to the employees is even more straightforwardly wrong because it's basically the same as the company selling the shares and giving the proceeds to the employee. It would seem that in their attempt to beautify Amazon.com's horrid earnings, analysts are reaching for murkier and murkier practices.