(Commentary by David F. Larcker, James Irvin Miller Professor of Accounting, Director of Stanford Graduate School of Business Corporate Governance Research Program, And Brian Tayan, MBA ’03, Stanford GSB; Date: 9-23-2010)
A new paper by Barth, Gow, and Taylor (2010) calls into question the practice of excluding stock options from reported earnings per share (so-called “non-GAAP” earnings).
Under U.S. accounting rules, companies are permitted to report adjusted financial results, so long as they disclose at the same time the most directly comparable GAAP figure and a table that reconciles the two. Experts have long clashed over the value of these adjustments. On the one hand, adjustments may be done to improve earnings quality when one-time items distort the long-term profitability of a company. In this case, excluding these items would improve the information quality of the reported figures. On the other hand, earnings adjustments may be made opportunistically by management to boost short-term results and beat Wall Street estimates. In this case, excluding certain items reduces earnings quality by misrepresenting the company’s underlying profitability. The evidence suggests that both can occur, depending on the circumstances.
In their paper, Barth, Gow, and Taylor (2010) examine the practice of excluding stock option expenses from quarterly earnings. They find that this practice does not improve earnings quality. Instead, they find that excluding stock options is done opportunistically “to increase earnings, smooth earnings, and meet earnings benchmarks.”
Technology companies tend to be the most enthusiastic practitioners of stock option non-expensing. This may be because they rely heavily on options in employee compensation packages, particularly when the companies are still in high-growth phase. For example, in Q2 2010, Google reported non-GAAP EPS of $6.45 compared with GAAP EPS of $5.71. VMware reported non-GAAP EPS of $0.34 versus $0.18 GAAP EPS. And Salesforce.com reported non-GAAP EPS of $0.29 versus $0.11 GAAP EPS. These are not small differences.
How does such a practice persist, particularly in a circumstance where it is clearly shown to reduce earnings quality? Isn’t this what governance systems are expected to prevent? And yet with stock option expensing, it seems that many members of the system are complicit: the board that approves the exclusion, investors who value the company on that basis, analysts who provide the research, and regulators who write the rules.
See also: David F. Larcker and Brian Tayan, “Pro Forma Earnings: What’s Wrong with GAAP?” CGRP-09.