Minding the GAAP: companies are using a wider range of accounting metrics to inform investors—and sometimes mislead them

In 1996, only 59% of examined companies showed non-GAAP metrics in their earnings reports. By 2018, that grew to 97%.

Comparisons of GAAP and pro forma results can vary wildly. weesen photos

If you were going to invest in a stock, which sounds like the better bet: a company where earnings per share went up 17 cents or a company where they went down $1.10? How about both at the same time?

The third quarter earnings of energy company Entergy in October 2019 showed exactly that. By one measure—standard generally accepted accounting principles, or GAAP—there was a big loss. Filtering out the contribution of a business line the company sold off and presenting non-GAAP, often called pro forma, results showed the improvement and how the remaining ongoing operations were doing.

Both ways of framing the data were valuable—that time. But, often, management teams use non-GAAP accounting in addition to GAAP to make their results look better. Just before WeWork gave up on its IPO, the company was at odds with the Securities and Exchange Commission, as the Wall Street Journal reported, over such things as how a novel metric, called “contribution margin,” tried to manage perception of heavy losses. (WeWork said it would “politely decline” to comment.)

The use of non-standard reporting—sometimes for a good reason and other times for marketing—has grown so much that all investors must keep a clear eye out.

“[Companies] use non-GAAP numbers to avoid reporting a loss or they want to avoid missing an analyst forecast,” says Carol Marquardt, a professor of accountancy and department chair at the Zicklin School of Business, Baruch College. “It is a mix of two motivations: to better inform investors and sometimes to mislead investors. And it’s difficult to tell which one predominates.”

Non-GAAP’s popularity

The practice of offering non-GAAP results has grown, according to data from Audit Analytics. In 1996, only 59% of examined companies showed non-GAAP metrics in their earnings reports. By 2018, that grew to 97%.

All financial reporting in the U.S. must be done using the standard generally accepted accounting principles, or GAAP, so investors can understand a company’s performance. However, companies can provide non-GAAP results, also called pro-forma, in addition.

“The biggest reason companies give for non-GAAP metrics is it’s more of what management sees as important to the company in the financial statement,” says Derryck Coleman, Audit Analytics research manager. Because GAAP accounting must apply to all companies, it often lacks specificity in a given industry.

For example, “gross revenues” (a measure included in GAAP) for hotels leaves much to be desired. “Occupancy [alone] doesn’t tell me much,” says David Larsen, fellow at the Duff & Phelps Institute, a think tank focused transparency and ethics. “I need to know the average revenue. If I say I’m running 80% occupancy but the reason I am is because I’m only charging $10 per room, [there is a problem].”

“There are some times when the GAAP accounting doesn’t make the most sense, like the amortization of goodwill,” adds Sandy Peters, a CPA and head of financial reporting policy at CFA Institute, an association of investment professionals. Goodwill accounts for the value of a company’s intangible property, including customer loyalty, patents and trademarks, brand, and employee relations. Amortization is the process of regularly reducing the value of property over its useful life.

Amortizing goodwill shows the lower value it has over time. But what is the lifespan of a brand? Should “Coca-Cola” be worthless now? Does a patent granted for 20 years still have significance when superseded by a newer technology?

“I think the accounting profession is lagging in many respects for business models that are coming out these days, like monetizing data,” says George Calhoun, an industry professor in the School of Business at Stevens Institute of Technology and director of the Hanlon Financial Systems Center. “I think there are legitimate moments when company can say the GAAP number is not showing the most accurate picture.”

Trying to follow the money

Multiple factors have contributed to perceived growth of non-GAAP reporting, according to Coleman. Mergers and acquisitions have expanded, complicating the view of how the continuing parts of a company are performing. Increased regulation around non-GAAP may also have just made such reporting more obvious to investors, even if many companies already used it.

Non-GAAP is perhaps most widely used to back out one-time expenses or revenues that aren’t representative of the company’s ongoing performance. This is where huge shifts in results can occur.

Johnson & Johnson, in its 2018 annual earnings report, showed nearly $15.3 billion net income under GAAP numbers but also offered a non-GAAP treatment that said net income was $22.3 billion after backing out a number of expenses, according to the company’s reconciliation document. Many of the categories—billions in total—also happened the previous three years as well to different degrees. (The company did not respond to a Fortune request for comment.)

Financial data company Calcbench ran an analysis on the S&P 500 in 2018 for Fortune. While there are many types of non-GAAP numbers companies can report, an important one is non-GAAP net income, which 256 firms used last year.

“The average non-GAAP net income for the S&P 500 is $2.46 billion,” says Calcbench CEO Pranav Ghai. That compares to an average GAAP net income across all the 500 of $2.38 billion, so not a large difference. However, compare the GAAP and non-GAAP figures for the companies that report both, and the results are dramatic. The average GAAP net income is now $2.06 billion. “The [change] is almost 20%,” Ghai says.

“According to research, on average, people tend to window dress [with non-GAAP] for the better and not for the worse,” says Gans Narayanamoorthy, associate professor of accounting at Tulane University’s A. B. Freeman School of Business. “This was very big during the dot com era. If you go back to 2001, the difference between pro forma earnings and GAAP earnings was huge.”

One study out of the University of Illinois at Urbana-Champaign, the University of Georgia, and the University of Arizona suggested that managers “exclude higher levels of both recurring and nonrecurring expenses in calculating the pro forma earnings number.” That would make earnings look higher. Also, managers “emphasize the pro forma figure by placing it more prominently within the earnings press release.”

There are clues to when the corporate intent is to sell more than inform. For example, a company might exclude what is a recurring item or the non-GAAP results are significantly higher than earnings estimates according to the Institutional Brokers’ Estimate System (IBES), an earnings estimates database owned by Thompson Reuters. “It tends to suggest that managers are reporting opportunistically,”

Which might be a sign that, at least for investors, they should mind the GAAP.

More must-read stories from Fortune:

—The stock market has hit 19 new highs in 2019 alone. Why?
—The 2020 tax brackets are out. What is your rate?
—How “VSCO Girls” are killing makeup sales
—What is “quantitative easing”—and why is everyone so worked up about it?
—What powered such a great decade for stocks? This formula explains it all
Don’t miss the daily Term Sheet, Fortune’s newsletter on deals and dealmakers.