SAN DIEGO (TheStreet) -- Say what you will about Amazon (AMZN) the stock, but you can't take this away from the online retailer: At least it tells you, right up front -- in the third paragraph of its earnings releases -- what its true share count is.
From its most recent earnings report, Amazon says:
"Common shares outstanding plus shares underlying stock-based awards totaled 475 million on September 30, 2013, compared with 469 million one year ago."
The key phrase is "plus shares underlying stock-based awards." Spelling it out, like Amazon does, gives analysts and investors an easy way to calculate the company's true market value and (if the company is making money) fully diluted earnings (or loss) per share.
The difference between diluted and the "basic" share count, as it's called, is often not that great.
The difference is beyond startling, with some dilution well in the double-digits.
Consider, for example:
- Yelp's stated number of shares is 65.5 million. Add in stock awards, and total shares are 76.6 million. The difference: 17%.
- Zillow: 36.7 million shares; 5,600 stock awards; total, 45,800. Difference: 25%.
- Pandora: 184.6 million shares; 34,176 stock awards; total, 218,843. Difference: 19%.
- LinkedIn: 113,940 Class A Shares; 23,971 Class B shares, 5,967 stock awards; total, 137,911. Difference, 21%.
- Twitter: Good luck trying to decipher the exact number from its IPO prospectus, but the difference between the headline share count and the real share count is around 25%.
Here's the rub: You wouldn't know it from the market valuations used by analysts from the big Wall Street firms, because they tend to just use the basic share count.
Why? My guess is, they use the same market-value numbers shown on Bloomberg, Factset or some other data service, which automatically mine them from the "basic" share-count line in SEC filings.
When companies are losing money, the basic and diluted share count is the same. As Yelp and other companies say in their SEC filings, as required by accounting rules, they intentionally exclude stock awards from earnings per share calculations, because "their effect would have been anti-dilutive for the periods presented."
In other words, it would make losses seem less than they really are.
Trouble is, that shouldn't matter to analysts when it comes to valuations and models. I would argue their models should be based on the full share count, even if they have to dig through SEC filings to find it.
Using anything less makes a mockery of their models. And in no small way, given the difference between the headline share count and the real share count. We're talking substantial dilution if these companies ever make money; in which case, earnings will be calculated off the real share count.
That's another way of saying: Current models on many of these companies, by many analysts, are steeped in the land of make believe.
Reality: If the analysts did their models the right way, using the real number of shares outstanding, whatever targets they have on the stock based on a multiple of revenues would be dramatically lower. Wouldn't want that in a market like this, with investment banking fees at stake, would we? (Oh, I forgot, that doesn't happen anymore.) All of this, of course, drives serious investors nuts, because they're looking at the total picture. As Brad Ginesin of Polar Capital says, "The playing field is not level." On one hand that gives the good analysts a leg up on one of the market's last great inefficiencies; on the other, like it or not, the Street tends to be pegged to the targets and estimates from analysts at the big firms. The solution? I propose the Amazon Rule. This would require all companies, at the top of their earnings releases, to state what their real share count is. This may not seem like a big deal now, but the numbers are the numbers. At some point the real numbers do matter.
-- Written by Herb Greenberg in San Diego