When it comes to the stock market, there is no single greater observable divergence right now than that between GAAP and non-GAAP earnings. As the chart below shows, while on a non-GAAP basis earnings have been hurting in recent years, with the LTM EPS of the S&P has declined to 116.4, down from 118.1 as of December 31, 2014, the real surprise is in GAAP EPS, which are back down to 88, a level last seen in 2007 when the market was about 500 points lower.
And, depending on whether one believes in adjusting EPS and giving companies the full credit of addbacks, pro forma numbers, and other various fudges, the P/E of the S&P as of this moment, is either 18x on a non-GAAP basis, or a ludicrous 23.7x if GAAP earnings are used.
The reason for the use of this non-GAAP numbers is very simple: to create an illusion of corporate profitability where one does not exist, something which can be seen most vividly in the earnings, or rather loss numbers of Alcoa, easily the most chronic offender of non-GAAP adjustments, which has made a half a billion dollar loss into half a billion dollar profit.
But, as the WSJ calculates, it is not just shareholders who enjoy the illusion that their investment is doing better than in reality (and thus their stock is worth more than it would be under a GAAP world) - it is corporate executives themselves who are delighted by non-GAAP numbers.
As the WSJ reports, non-GAAP adjusted metrics aren’t just showing up in earnings releases: "pro forma figures have been proliferating in annual proxy statements, too." And the reason why in addition to shareholders, management is also infatuated with non-GAAP is because "when used with compensation metrics, they can help executives draw bigger pay packets."
In other words, in the worst possibly form of skewed incentives, CEOs are ever more compensated to fabricate the most inaccurate profit picture they can come up with.
According to research firm Audit Analytics, the term “non-GAAP” appeared in 58% of proxies for companies in the S&P 500 that have released them so far this year. Five years ago, that term showed up in 27% of proxies for current S&P 500 constituents.
And this is how a major conflict of interest has emerged:
There is nothing improper about using non-GAAP measures as long as they are disclosed properly. And corporate boards decide on the measures they want to use for compensation purposes. Plus, there is an argument to be made for sometimes excluding items from results for compensation purposes. If, say, a natural disaster hits a company with expensive repairs, perhaps an adjustment is in order.
But other items that often get excluded in pro forma results, such as layoff-related charges, do seem like a reflection of management’s performance. And boards have too often shown a willingness to set awfully low bars for executives to clear.
Why conflict of interest? Because management teams are progressively rewarded to come up with untrue, borderline fraudulent numbers. After all, why else is the SEC allegedly "cracking down" on companies to limit the number of non-GAAP adjustments. As the WSJ itself admits, this push to rewarding management based on non-GAAP numbers disadvantages shareholders and wrecks the idea of pay for performance. "In that vein, the dramatic rise in the number of companies using pro forma measures to determine bonuses would indicate the balance between shareholders and executives is being skewed in executives’ favor."
Some Dow Jones case studies, where as the WSJ notes an examination of the most recent proxy statements from companies in the Dow Jones Industrial Average shows about a dozen of the index’s 30 constituents had annual pro forma earnings well in excess of GAAP ones and used the pro forma ones in annual bonus calculations.
Coca-Cola’s pretax income increased by 3% under GAAP. But after adjusting for the impact of the stronger dollar and “nonrecurring items,” the income growth figure the Dow component used for bonus purposes rose to 5.5%.
Absent adjustments, Dow member Home Depot reported operating income of $11.77 billion last year. But the pro forma figure of $12.06 billion it used for compensation figures excluded the impact of the strong dollar and costs associated with its 2014 credit-card data breach.
Pfizer earned $1.11 per share last year under GAAP. But the Dow component excluded a number of items from the pro forma results it used for bonus purposes, pushing earnings per share to $2.20—above the $2.05 it had targeted for its annual incentive program. Without that adjustment, the chief executive’s bonus under the company’s annual incentive program could have been significantly lower. While in most years, the pro forma earnings the pharmaceutical company has used for compensation purposes have exceeded GAAP, in 2013 they were lower.
And if the WSJ thinks the DJIA is bad, please don't even think of looking at the NASDAQ and/or the Russell 2000, where the only difference between profit and loss is the FASB and SEC's overly generous treatment of non-GAAP reporting.
The biggest risk here, however, is a simple one, and the WSJ explains it well:
"more worrisome, using pro forma to set bonuses provides executives with an incentive to exclude items not because they should, but to hit performance bogeys. That creates a risk that pro forma results say less about a company’s underlying health than about executives desire to get paid more."
To be sure, we have been saying precisely this since 2010. We can only hope that now that the WSJ has also caught on, that the SEC will finally do something about it.
There is a problem, however: a full crackdown on non-GAAP would mean that instead of generating 116 in EPS, the S&P500 actually made 25% less in profits. Which also means that if one assumes an 18x multiple, the fair value of the S&P is just under 1,600.
So the question is whether the SEC would ever willingly step into a mine field where the crackdown on fabricated earnings would mean wiping out trillions in fabricated market cap from the S&P500. The answer: of course not. Which is why while the SEC may rage all it wants against non-GAAP adjustments, it will never actually do anything.