Non-GAAP Measures and One-Time Adjustments

This October, PwC released a study on the use of non-GAAP metrics (NGMs) in IPO prospectus filings. According to the study, almost 60% of over 400 companies used at least one NGM. Out of those, at least 65% used EBITDA and/or Adjusted EBITDA (the latter amounts to 46% of the overall population). NGMs are often used to exclude certain events that may affect the comparability of financial statements.

Interest in the use of NGMs was boosted not long ago by Groupon’s controversial use of non-GAAP in their IPO prospectus. Back in 2011, Groupon presented Adjusted Consolidated Segment Operating Income (CSOI) that eliminated certain marketing expenses. The SEC objected to this presentation and Groupon ultimately removed the metric.

In this blog we focus on a slightly different aspect of NGMs – namely, whether extensive use of non-GAAP language in financial reporting documents could be used as a leading indicator of subsequent one-time adjustments. Numerous non-recurring adjustments are often associated with lower quality of financial reporting. (We recently wrote about this with respect to GT Advanced Technologies here.)

We examined 8-K’s of Russell 3000 companies filed in Q3 of 2013. Out of 3,000 companies, more than 2,000 (about 67%) used ‘non-GAAP’ words at least once in a press release, while more than 550 companies ( about 18%) used the same kinds of words at least 20 times. We also checked the frequency of occurrence of one-time events, like changes in accounting estimates and out-of-period adjustments in the subsequent four consecutive quarters (from October 1st 2013 till September 30th 2014).
The results are presented in the following table.

Evidently, companies that place a heavy reliance on non-GAAP are more likely to disclose an out-of-period adjustment within a four-quarter time span. We ran the same analysis on the data from Q4 2012 and Q1 and Q2 2013 (not included in the table above) and came up with substantially similar results. Interestingly, those one-time adjustments were rarely (if ever) excluded from GAAP results to arrive at the cited NGMs.

We also checked whether the companies with substantial use of ‘non-GAAP’ language were more likely to have a restatement, and the initial results indicate that they are.

Table 2 copy

The table above shows that the frequency of restatements was also higher for companies with extensive non-GAAP language. It is important to note, however, that most of the restatements during this period of time were revision restatements that do not undermine reliance on previously filed financial statements, and are in general less material than non-reliance errors (i.e., those requiring an Item 4.02 8-K disclosure).

Of course, this analysis is only based on a small sample, covering just a handful of quarters worth of data. Also, changes in accounting estimates may include standard and recurring charges, such as those related to percentage of completion accounting. Of note is that non-GAAP language, while strongly correlated to NGMs, is not a perfect substitute. Some companies that extensively rely on NGMs may avoid using non-GAAP words by replacing them with “non-core” or “adjusted”. Further research may reveal more interesting indicators and/or correlations between various forms of non-GAAP metrics used and other financial and/or business events.