In Sunday's New York Times, respected columnist Gretchen Morgenson took aim at public companies' proliferating use of non-GAAP financials in a column titled "Fantasy Math Is Helping Companies Spin Losses Into Profits." This appears to be a concern beyond the Times; a number of SEC officials have highlighted the increase of late as well.
I'm not in the habit of publicly calling out Pulitzer Prize winners (or, in this day and age, regulatory agencies), but the context of Morgenson's piece, in particular, struck me as both highly overwrought and offensive to investors. The idea seems to be that GAAP principles keep companies from if not outright lying, then misleading their shareholders, with Morgenson writing that "the gulf between reality and make-believe in these companies' operations is so wide that it raises critical questions about whether investors truly understand the businesses they own."
This isn't a new concern; reporting issues gained attention from the SEC at the beginning of the last decade, in the wake of the Enron and WorldCom scandals. That led to Regulation G, which requires the now-standard "reconciliation" of non-GAAP figures at the end of earnings releases. With usage seeming to increase - Morgenson cited a study that said 90% of S&P 500 companies reported non-GAAP figures in 2015, up from 72% in 2009 - the issue is back at the forefront.
In my mind, non-GAAP earnings really aren't that big a problem. Regulation G was a common-sense, and effective, solution to the issue: if companies are going to exclude (or include) revenue or costs from non-GAAP figures, they need to do so clearly and specifically. With that reconciliation, non-GAAP figures strike me as a benefit to investors, if only for the very simple reason that they provide more information. Is that information sometimes used by management to cast operations in a more favorable light? No doubt. But even the decisions on which factors to exclude/include themselves can provide information for investors relative to the trustworthiness of management.
There are just a host of problems with Morgenson's argument, to be blunt. In fact, I'd argue that assuming non-GAAP reporting to be "phony-baloney," in her words, is dangerous, particularly for individual investors. Those figures are simply information, numbers, data; they can be manipulated or twisted like any other statistic. But they also offer value, and ignoring those figures or assuming they are somehow slanted misses out on some of the benefits they provide.
1. GAAP isn't the truth, the whole truth, and nothing but the truth.
In the title, Morgenson refers to non-GAAP figures as "fantasy math." In the article, they are "phony-baloney," "massaged," "make-believe," "made-up," "fake," "fantasy figures" and "make-believe" again.
The implication, then, is that non-GAAP numbers aren't real, unlike their GAAP counterparts which are set in stone. And Morgenson is supported by no less than former Enron CFO Andrew Fastow, who told students at Dartmouth in 2012:
I used loopholes in the rules to get around the principles of the rules.
Actually, Fastow was referring to GAAP rules, though Enron did report non-GAAP "recurring earnings" as well.
The point is that GAAP accounting isn't some sort of black and white system, whereby 1 + 1 = 2 while non-GAAP figures mean 1 + 1 = $50 million more in CEO compensation. GAAP's nature is defined in its title: Generally Accepted Accounting Principles. They're "generally accepted," not "iron-clad." Indeed, those principles change from year to year; Caterpillar (NYSE:CAT), for instance, will see a non-cash benefit of ~$0.50 in EPS this year due to a change in calculating pension costs.
That benefit comes under GAAP accounting - but, being non-cash, and non-recurring (the impact of those costs will vary in 2017 and beyond), couldn't one consider that "phony-baloney?" As an investor, what would you pay for $0.50 per share in one-time, non-cash earnings? At $78, should CAT's price-to-earnings multiple be considered 24.4x (excluding the $0.50 one-time non-cash charge) or 21.1x (at guidance for $3.70)? (If you follow CAT closely, you know those figures exclude restructuring costs as well and are thus non-GAAP, but for simplicity's sake let's pretend otherwise for the moment.) If EPS grows to $4 next year, did profits improve by 8% (over $3.70) or by 25% (over $3.20)? The answer isn't as simple as, "well $3.70 is the GAAP figure." As an investor, $3.20 is the more important figure to me; regardless, using $3.20 at a base isn't "massaged," "fake" or magic. It's an adjustment.
I've had this argument before, in comment streams and offline, but the idea that non-GAAP is inherently misleading thus implies that GAAP accounting is simple and straightforward. Ask an accountant; it's not. There are different ways to understand revenue, costs and profits, and myriad different ways to extrapolate what those trends mean for the discounted sum of future cash flow - which is what shareholders are purchasing when they buy equity in a company.
What's dangerous for individual investors, in particular, is that the idea that "GAAP good, non-GAAP bad" can then be taken to mean that there's really only one way to value earnings - and thus a stock pricing. That's patently untrue, of course. The difficulty of stock-picking is not "phony-baloney" non-GAAP earnings; the difficulty is the massive amount of information available, the huge pool of other participants trying to find their own alpha, and the essential impossibility of predicting the future. (Those factors also drive all the fun of stock-picking.) It's a short step from "these are the only numbers that are true" to 'rules' like "a P/E over 50x is too much, and a P/E under 10x is good" or "a 4% dividend yield is good" or "debt is bad." And those types of simplifications can be very, very dangerous for individual investors - far more so than a little bit of shine on some non-GAAP earnings figures.
2. Management will always, or often, cherry pick the figures.
This seems to be the implied point Morgenson makes when writing that the 90% penetration (so to speak) of non-GAAP in the S&P 500 means "a modest problem [has grown] into a mammoth one."
That's not always the case. In my experience, companies exclude benefits, such as the reversal of the valuation allowance or gain on sales. That said, the adjustments usually benefit reported earnings, in particular. Morgenson points out that of 380 companies studied, non-GAAP figures showed a 6.6% year-over-year increase in 2015 earnings; GAAP results, however, implied a decline of almost 11 percent.
Aha! This shows that those dastardly corporate executives are tricking us poor investors. Instead of posting any growth at all, the companies actually are declining. At this rate, we'll all be broke in a matter of years and it's all because of those @#$#@! non-GAAP numbers.
There's another explanation, though: accounting. The majority of events excluded from non-GAAP figures are going to be negative, particularly when it comes to non-cash charges. Ever heard of a goodwill write-up? Other than in M&A accounting, it doesn't exist. In fact, according to GAAP, "under no circumstances can a company write up its fixed assets even if the market value of these assets exceeds its book value." Companies who make poor acquisitions and have to write down goodwill are going to exclude those non-cash charges; companies who make accretive acquisitions don't have the option of excluding non-cash gains, since they don't exist. Looking at non-cash charges - a very common type of adjustment, and the most likely to be material - they are almost always going to be negative. Thus revisions on average will improve earnings, no matter the intent of CFOs and CEOs.
But that doesn't mean non-GAAP figures are inherently untrustworthy, or that "trick" accounting is being used to cover earnings declines. The 2015 figures likely were impacted significantly by oil and gas companies (as Morgenson alludes to in the piece), and in general non-GAAP figures are going to look better than GAAP figures. That doesn't mean they're "fake," or that executives are misleading their shareholders.
3. Investors are too dumb to understand.
Morgenson's claim that non-GAAP usage "raises questions about whether investors truly understand the businesses that they own" irritates me, to be blunt. First off, the GAAP numbers are right there. If an investor wants to say, "I don't trust non-GAAP figures," he or she can skip them. Yes, conference call discussions and the like tend to be non-GAAP-heavy, but there's also a number of companies - 50 in the S&P 500 alone, apparently - that don't bother with non-GAAP. It's investors' responsibility to do their due diligence; barring outright fraud (not part of the discussion here), a failure to understand a business is a failure on the part of the shareholder, not management.
Secondly, Regulation G requires a reconciliation of non-GAAP to GAAP - and that in and of itself can be very valuable. In 2014, on this site, Jay Yoon pointed out that VOXX International (NASDAQ:VOXX) was reporting its non-GAAP net income without considering the tax implications; in other words, it was taking its adjustments, adding them 100% to GAAP net income, and using that figure as non-GAAP EPS. That red flag led Yoon to review management statements in more detail; that article colored my view of the company when I looked at it as a potential buying opportunity, and continuing management concerns led Yoon to make a profit and me to avoid a loss on a stock that has collapsed. Conversely, Darren McCammon, who has been dead right in recommending Eldorado Resorts (NASDAQ:ERI) on this site, pointed out the lack of a pro forma adjustment by the company in supporting his case for shareholder-friendly and conservative management.
Some might argue that giving management rope with which to hang itself isn't necessarily a point in favor of using non-GAAP numbers, but I disagree. There's value in seeing what exactly management decides is to be excluded, what is "one-time," and to what extent they are trying to shape their own numbers. When companies have restructuring costs in nine out of ten years, or have stock-based compensation that is a large majority of operating income, it doesn't mean non-GAAP numbers are per se wrong. It's actually a piece of information - and often a very helpful piece of information - as to the priorities of management.
As long as the numbers are truthful, and reported under Regulation G, non-GAAP simply provides another way of looking at the financials. Again - another way, not a "make-believe" way. On the balance, it's likely that non-GAAP figures are presented in a way that makes management look flattering. But I'd also wager the obesity crisis in the U.S. would look a lot less pressing if a researcher used the weight we list on our driver's licenses.
Meanwhile, assuming that investors will get swayed by non-GAAP figures into making the 'wrong' decision assumes that investors don't do their homework. I don't think that's the case - and, if it is, I don't feel any sympathy for investors who simply listen to management. That's true beyond the numbers as well: according to management, every corporate initiative is going to go well, is going well, and would have gone well except for macro/weather/act of God/stupid government people/bad luck/the guy we fired. The SEC has many, many more pressing concerns than protecting investors from themselves - as long as the information is readily accessible. Yet there's an SEC official complaining in public that CNBC reports non-GAAP figures without pointing out the adjustments. Heck, no one watches CNBC anymore, anyways.
A very interesting example of this is Zynga (NASDAQ:ZNGA), where I've argued for years that the company is massively unprofitable despite posting positive Adjusted EBITDA, since it excludes share-based compensation from that figure. In 2015, for example, share-based comp was $132 million; Adjusted EBITDA was $17 million.
Should ZNGA not be allowed to report that Adjusted EBITDA figure, or mention it on conference calls? Would it protect some ZNGA investors from buying into a company that some erroneously think is making money? After all, some of those investors perhaps are influenced by Chairman/again-former CEO Mark Pincus using the term "profitable" in his discussion of results - when accounting for share issuance, the company is nowhere close. (Of course, calculating share-based compensation is itself a bit of a gray area, as models and inputs impact what the actual present value of that compensation is. Accounting isn't easy.)
Personally, I prefer the non-GAAP results because it gives me a clue as to what Pincus' motivations are. The fact that Zynga so carelessly issues stock to employees (and many, many former employees) is a valuable piece of information to understand when valuing the stock. The chairman and controlling owner of the company doesn't seem to care about dilution. To me, that is a useful fact. Would Pincus care more if he had to admit that the company was losing over $100 million a year? Would he perhaps treat his shareholders better? Maybe. But the information, again, is right there, right at the end of the press release - and I think it's better to know Pincus' attitude than to have it buried amidst requirements on how the company discloses its numbers.
The answer to potential misinformation, in my opinion, is not less information. That argument is based on the argument discussed point #1, that GAAP and non-GAAP somehow are "right" and "wrong," "true" and "false." They're not. They're different.
4. GAAP accounting sucks.
Below is Chesapeake Energy's (NYSE:CHK) earnings per share over the past five years:
source: CHK 10-K
What is an investor supposed to do with those figures? Of what value is EPS of -$22.43? On a GAAP basis the stock trades below $7, meaning its trailing P/E is about -0.3x. Doesn't that sound cheap?
In all seriousness, the 2015 EPS number is essentially useless: it is full of non-cash charges, including a $13.976 billion non-cash impairment of the company's properties. Is that figure of any real value? Are there investors who would look at the stock and think, "Well, if the impairment were under $10 billion, I'd probably buy shares - but $14 billion is just too much." Is the market surprised that Chesapeake had to take those types of write-downs this year? Of course not; oil crashed, shares fell and then the company wrote down its assets. It wasn't the other way around; the write-down isn't driving the operational weakness, meaning its exclusion in the calculation of adjusted net loss ($0.20 per share) is 'hiding' some dark truth about CHK's earnings.
Most of these nefarious non-GAAP exclusions, truthfully, are just things that get in the way of the important goal of trying to understand a company's true earnings power. Very few investors I know care what goodwill is, or value a stock based on a multiple of book value. Sometimes those figures have relevance: if a company is writing down goodwill every three quarters, its acquisition strategy probably isn't working out. If a company is trading below book value, that could imply a potential 'floor' to the stock, but most good investors only consider tangible book value, which ignores the non-cash intangible assets created by... GAAP accounting. (Those intangible assets always struck me as the definition of 'make-believe,' for what it's worth.)
I see it as much more difficult to analyze a company that doesn't use non-GAAP accounting. I'd much rather have the reconciliation and the one-time impacts called out than have to comb through a K trying to figure out what property impairments were or understand the amortization of intangible assets from a 2013 acquisition ran through the P&L in Q3. I know the non-GAAP figures often will exclude everything possible, but at least those figures are listed and from there I can make my own adjustments, if I think management is being generous.
For instance, I usually add back share-based compensation, in an attempt to calculate net income growth. But analysts who are using an EPS-based model can account for that compensation by using non-GAAP numbers as a base and then including dilution going forward.
At the end of the day, that's kind of the point: stock analysis isn't black and white. It's not a science; it's an art. Or it's just monkeys throwing darts. Accounting isn't black or white, either, one reason why assuming non-GAAP figures are "fantasy" is far too simplistic.
That assumption also impugns corporate executives; there are multiple examples of non-GAAP figures being worse than GAAP figures (here's The Container Store's (NYSE:TCS) Q4 release, where FY14 adjusted net income was lower than reported, thanks to a one-time gain on sale, and FY15 net income wasn't changed despite a number of accelerated investments in the business). And it impugns investors, in particular individual investors, who apparently can't see non-GAAP figures on CNBC without rushing out to buy shares of Crapola Inc.
The larger problem is that the world just isn't that simple. Corporate executives aren't all bad; investors aren't all dumb; accounting isn't easy and neither is stock-picking. There isn't a 'true' figure for how much a company earns any more there is a 'true' fair value for its stock. All investors can do is get as much information as possible - while understanding the source. The idea that some of the information is 'right' and some 'wrong,' to me, is more make-believe than anything corporate executives can come up with.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.