Earnings do not reflect how much employee stock compensation costs shareholders
The COVID-19 pandemic and attempted government insurrections aside, headlines these days are dominated by how Big Tech’s growth has been subsidized by the section 230 exception of the Communications Decency Act. And while section 230 protects websites from lawsuits in case a user posts something illegal, I want to focus on another concession enjoyed by Big Tech. It’s an arcane, but effective financial reporting rule that allows companies to under-report the true cost of compensating their workers.
So how is this possible? Tech companies rely on generous equity grants to attract and retain employees. Reporting rules allow these companies to expense the projected fair values of these shares and options on the date at which they’re granted. The actual cash flow associated with these grants can be higher relative to the compensation expense that shows up in companies’ income statements. However, the cash outflow is obscured as payouts to shareholders are never explicitly linked back to compensation expense. Tech companies also feel the pressure to turn a profit to assure investors that they will grow into their lofty valuations.
For example, consider the case of Netflix. I extracted the following footnote related to stock options from their 10K filed for the year ending on December 31, 2019.
Consider the 2.208 million options exercised by employees in 2019. These options were issued at an exercise price of $32.88. When employees exercise these options, Netflix needs to issue share certificates to these employees. Where do these shares come from? Many cash rich companies such as Google would buy stock from the open market to satisfy that demand. Netflix needs all the cash it can get to invest in content production. So, it issues equity to meet the demand for share certificates to its employees. The average share price at which Netflix traded in 2019 was $320.66 (assuming it is the same as the exercise price at which options were granted in 2019). Hence, the shareholder value given up by Netflix in 2019 to pay its employees was $320.66*2.208 million shares or $708 million.
How much of this value transfer to employees is recognized in Netflix’s income statement as compensation expense? That is somewhat difficult to nail down. However, we do know that the “fair value” of these options on the grant date of the option is required to be expensed as per U.S. GAAP (generally accepted accounting principles). In general, the “fair value” of the option, given Netflix’s profile, is roughly half or a third of the exercise price of the option. To help the argument against me, let’s go with half. That would imply that previous income statements of Netflix were charged with compensation expense of approximately $36 million (2.208 million shares*$32.88/2). Forgetting taxes for a minute, compensation expense is under-reported by $672 million ($708-$36) relative to the value sacrificed by Netflix’s shareholders to its employees.
Before you think this is insignificant, let’s benchmark the unreported expense to previous year’s income numbers as these options were likely issued dating back to 2017, 2016, or even earlier. The pre-tax income for Netflix in 2017 was $485 million. The unreported compensation expense thus works out to 139% of its 2017 reported income ($672/485 million). Hence, Netflix would have reported a loss in 2017 had the true cost of employee compensation been reported to shareholders.
There are countervailing arguments for sure. One could argue that the options were given back in 2016 or 2017 when the future run-up in stock price was not expected. Fair enough. But we routinely mark other financial instruments to market. Why should firms take write downs for marketable securities that have lost value?
The second objection I often hear is that giving employees options back in 2016 or 2017 actually enabled Netflix to do well and led to the massive run-up of its stock price. That is certainly possible. Assuming stock prices are at least loosely correlated with realized performance, Netflix’s success will show up as higher revenues in its income statement. Instead, why not reflect the increased costs of compensation potentially responsible for such success? Moreover, would the stock have run up as much as it did if investors realized the actual value transferred to its employees? I am not sure, but it would be good to have these questions answered in practice.