Analysts have been widely inconsistent by accepting the use of some company's non-GAAP earnings per share accounting for estimates, such as eBay (EBAY) and Google (GOOG), and then holding other companies such as Apple (AAPL) and Amazon (AMZN) to a higher standard by using straight GAAP earnings for estimates. Using eBay as an example, their CY13 non-GAAP EPS estimate is 20% higher than the GAAP estimate. This translates into a 17% lower forward P/E estimate (currently 18.5 vs. 22.2) on all financial websites and discussions about the stock by writers and reporters. This 20% inflation of earnings is highly misleading and can lead to poor investment decisions/comparisons if the investor does not know the estimate is non-GAAP or what non-GAAP excludes vs. other securities.
Non-GAAP (also to referred to as pro-forma or adjusted results) reporting is not new and has been utilized by companies for a long time to remove large one-time items (like restructuring, discontinued operations, acquisition expenses, etc..) that skew GAAP results from quarter to quarter in order for the investor to better track the on-going operations of the company. It was never meant to be a continuous operating reporting result. To compound matters, each company that reports non-GAAP earnings can exclude different expenses, as non-GAAP is defined by the individual company. That means that the if you want to compare non-GAAP results of Google and eBay for example, they may still be apples to oranges. One expense that is almost always excluded in non-GAAP results is employee stock compensation expenses.
Non-GAAP results became very popular for Tech companies during 2006 as FAS 123((((R)))) became effective on the start of a public company's fiscal year after June 15, 2005. FAS 123((((R)))) outlined the SEC's new stock compensation expensing requirements and had a major impact on companies that utilized stock options and RSU's as a significant portion of their employee's overall compensation. Tech companies fell right into the sweet spot of impacted companies due to the practice in Silicon Valley of using piles of stock to attract and retain key talent. Based on the significant discontinuity in earnings FAS 123((((R)))) created, many companies reported pro-forma results initially to help analysts understand the impact of the new rule.
FAS 123((((R)))) essentially requires companies to estimate the potential total compensation an employee will receive from the stock grant and then amortize quarterly that non-cash cost as an expense (either in cost of sales or operating expenses) over the lifetime of the grant. While I have my thoughts on stock compensation expensing, this article is not meant to take a position on it one-way or another, but to point out the large inconsistency in analyst estimates as they include it for some companies and not for others.
Figure 1: eBay Non-GAAP Expense Exclusions
Figure 1 displays what costs eBay excludes in its non-GAAP reporting. There are essentially two main exclusions: stock-based compensation and amortization of acquired intangible assets. As you can see, neither of these expenses is a one-time expense - they are both recurring and significant. They both hold in common that they are non-cash charges, but both do have impacts on cash position and/or dilution. Stock compensation either increases dilution or requires the use cash to buy back the shares. Even if the shares are not bought back, I would suggest stock compensation is a real cost, as the company could have issued that stock to the public and retained the cash rather than pay it to the employee. Amortization of acquired intangible assets is the result of paying a premium for a company (or Intellectual property) with either stock or cash, so this is a catch up charge that gets expensed over the useful life of the assets so the company does not get a large charge at the time of acquisition.
While Amazon details the expected stock compensation charge in their quarterly guidance, they do not provide a non-GAAP table and analysts do not back it out of their earnings projections. While I have been extremely critical of Amazon's business plan and transparency, they at least focus and report consistent with GAAP. If Amazon had similar non-GAAP reporting to eBay, Analysts would have had them earn about $950 million in 2012 instead of reporting a $40 million loss. No one can tell me eBay's stock compensation expense is less relevant than that of Amazon or Apple.
In my research, it appears clear that analysts simply follow the company's reporting preference. If the company focuses on GAAP, analysts use GAAP, if they focus on non-GAAP, analysts use non-GAAP. It has been eight years since stock expensing went into effect, and there is no longer a need to show the difference in earnings with and without stock expensing. It is readily available in the cash flow statement. We should be able to expect a level of consistency within an industry, but apparently that is asking too much.
Analysts typically follow a company's lead in what costs they include in their earnings forecasts rather than maintain a consistent approach to earnings forecasts across all companies within an industry. This leads to misleading comparisons of company valuations due to some stocks having inflated past and forward EPS forecasts.
Additional disclosure: I am long AAPL.