A discussion with Fletch about GAAP and non-GAAP

Investment Analysis Clubs Saul’s Investing Discussions
1

Fletch and I have always disagreed about this. Having an accounting background, he tends to prefer GAAP, while having a practical background, I prefer adjusted (although I believe he agrees that the repricing of warrants causes all kinds of distortions). We have had the following discussion in the past 24 hours that I thought people might enjoy reading. It’s posted here with Fletch’s concurrence.

I started off by asking him: “I just looked at CELG’s results. GAAP earnings for the quarter almost tripled. They were up 196% (from 25 cents to 74 cents). Non-GAAP earnings were up 33% (from 76 cents to $1.01). Which do YOU think gives a more realistic, more comparable, more consistent, more reproducible, view of the growth of the business and its real earnings in the quarter?”

Fletch responded

To give you a serious answer, as long as everyone does everything the same way, I don’t care what method of accounting anyone uses. My main concern has always been the ability to compare apples to apples between different companies. As long as every company measures non-GAAP earnings the exact same way, great - go ahead and get rid of GAAP. I really don’t care about the specific issues (expensing options, repricing warrants, etc.) as long as everyone presents their results using the same, consistent accounting. To use an admittedly extreme example, what if one of your company’s reported results based on accrual accounting and a competitor reported its results using the cash method? Without spending a tremendous amount of time, how on earth would you ever be able to compare the results against each other? So that’s what I care about - consistent accounting and reporting for all companies. I don’t really care about the specific rules themselves.

My response was:

“Interesting response, and I understand what you are saying about comparing across companies. I’d guess, though, that 90% of companies that report adjusted earnings actually do the same thing, not different things. They back out stock-based compensation and repricing of warrants, and occasional one-time events. (I know some people claim that companies claim one-time events every quarter, but if you look at reputable companies, that’s not really what they do.) What I like about the adjusted earnings are: First, they usually (usually, that is), reflect better what is really happening at the company, and Second, they are more comparable quarter to quarter in the same company (rather than a jerry-rigged GAAP system that raises or lowers reported earnings each quarter depending on whether the stock price is up or down).”

And Fletch responded:

This is also why I’m a big fan of the cash flow statement. In many ways, the information on the cash flow statement is more indicative of a company’s performance than EPS, adjusted or otherwise. There is a lot of room for management to manipulate earnings, even adjusted earnings, but manipulating the cash flow statement is more difficult, in my opinion.

And I sure wouldn’t argue against checking out the cash flow statement, either.

Saul

PS By the way, you’ll notice that Fletch avoided ever answering my initial question about Celgene’s earnings: “Which do YOU think gives a more realistic, more comparable, more consistent, more reproducible, view of the growth of the business and its real earnings in the quarter?” (It obviously wasn’t the GAAP earnings).

:wink:

6 Likes
2

Saul,

To answer your question, I would use the GAAP income statement combined with an analysis of the cash flow statement. I do not, and will never, trust non-GAAP or adjusted earnings numbers because management is free to do whatever it wants to those numbers, quarter after quarter, presenting them however management wants without any consistency (it should also be pointed out that non-GAAP numbers are not audited, so you don’t even know if management’s numbers are correct). Management can play fast and loose with adjusted earnings numbers and you’ll never have any way of knowing what it’s doing without digging through the minutia of every press release and earnings. Groupon, anyone? No thanks.

I think a review of GAAP earnings numbers in conjunction with a review of the balance sheet and cash flow statement will tell me everything I need to know about a company’s results. I have no need for non-GAAP earnings.

Fletch

14 Likes
3

By the way, you’ll notice that Fletch avoided ever answering my initial question about Celgene’s earnings: “Which do YOU think gives a more realistic, more comparable, more consistent, more reproducible, view of the growth of the business and its real earnings in the quarter?” (It obviously wasn’t the GAAP earnings).

Your point about warrants is a very good one Saul but where I have a problem with non-Gaap is when they back out stock compensation. Some companies give a large amount of stock options and I am wondering if it really is beneficial to give them a pass on that. What do you think or could you expound on this?

Also, I have a problem with companies that keep buying up other companies. It can be great for companies to acquire other companies, like Syna recently did, but when they become serial acquirers shouldn’t that be reflected as a business expense? What do you think?

Andy

4

where I have a problem with non-Gaap is when they back out stock compensation. Some companies give a large amount of stock options and I am wondering if it really is beneficial to give them a pass on that. What do you think or could you expound on this?

Andy,

EVERY company that gives adjusted earnings backs out stock-based compensation. That’s not just a coincidence. It’s because it’s double counted. If they give stock-based compensation, the extra stock is counted in the diluted shares, which reduces earnings per share. To double-count it also as an expense is silly.

Here’s an extreme example which will make it clear.

Company ABC has a 100 million shares of stock and is selling for $10 per share. (Capitalized at $1 billion). They made $50 million, or 50 cents for each of the 100 million shares (50 cents per share). Thus a PE of 20 which is very reasonable.

They decide the CEO did so well that they give him, as a one-time bonus, 5 million shares (which is worth $50 million). They now have 105 million shares, and their $50 million is divided by 105 million, giving them 47.6 cents a share in earnings instead of 50 cents.

Should we also subtract $50 million from earnings, giving them zero earnings, because they gave away that stock? That’s silly, as they have already deposited that $50 million to their bank account and can use it to build a factory, or make an acquisition, or whatever. It’s there! How can you say they had zero earnings? That’s the kind of meaningless calculation GAAP requires (I’m not an accountant, and my figures may be off, but you get the idea).

Best

Saul

4 Likes
5

where I have a problem with non-Gaap is when they back out stock compensation. Some companies give a large amount of stock options and I am wondering if it really is beneficial to give them a pass on that. What do you think or could you expound on this?

This is quite a complex issue. There can be no doubt that stock options are compensation, The issue is how to properly account for it. I’m not sure if my English terminology is entirely correct as I’m translating from Spanish.

A corporation is a “legal person” entirely separate from the stockholders who can be “legal” or “natural” persons. Not too long ago the Supreme Court allowed corporations to pay for political ads precisely because they are “persons.” It’s kind of weird that non-human persons should have the same prerrogatives as human persons. Additionally, legal persons do not own themselves, they are owned by stockholders.

If you don’t own yourself you cannot give yourself away. That would be stealing from the rightful owners. Stock option plans have to be approved by the Assembly of Stockholders. What this means from a legal and accounting point of view is that stock options are not given by the company but by the owners of the company. If you own a business you can give half to a friend but the business itself cannot give away half of your property right in the business to anyone.

By expensing stock options GAAP makes a mockery of this legal distinction.

A stock option is more like a secondary offering than like a payment, it creates dilution, not an expense.

Denny Schlesinger

5 Likes
6

A stock option is more like a secondary offering than like a payment, it creates dilution, not an expense.

What a great way to put it, Denny!

Saul

7

EVERY company that gives adjusted earnings backs out stock-based compensation. That’s not just a coincidence. It’s because it’s double counted. If they give stock-based compensation, the extra stock is counted in the diluted shares, which reduces earnings per share. To double-count it also as an expense is silly.

Saul, I have to give you credit, you’ve sucked me back into a board conversation I swore I’d never have again with you. What’s that Al Pacino line?

You’re right. Every company does back out stock-based comp when it delivers non-GAAP adjusted EPS. You know why? Because it makes management look better! Earnings with stock-based compensation excluded will always be higher than GAAP earnings, so management LOVES to give out that number and say, “See, we actually did even better than that dirty SEC says we did! Aren’t I an amazing CEO?”

The shares related to option grants are counted in diluted EPS, but if you don’t expense the shares, nothing shows up in operating income. So despite having awarded something that was cleary compensation (otherwise why would anyone want them?), if the cost of the stock options are not included in operating results the associated expense does not show up in the income statement, which gives one a distorted view of the company’s actual results. The options would purely have a balance sheet effect when exercised via dilution of shareholders equity.

Yes, the diluted EPS number already includes the options (so perhaps don’t bother with diluted EPS?), but that doesn’t help me if I’m trying to compare quarter over quarter SG&A growth, for example. If Company A compensates its executives with stock options, but Company B only pays its executives with cash, I would never be able to accurately compare the operating margins against each other unless the stock options were expensed as compensation. Without the value of the options included in an expense line item, Company A’s operating results would look far superior to Company B’s when in fact both companies have actually awarded their executives with pay packages of equal value. It is completely unfair that Company A would show a higher operating income than Company B simply because it paid its management team in options instead of cash.

Yes, the diluted share count would show the “true” EPS even if options were not expensed. But EPS does not help me examine margins, whether not SG&A expense is growing at a faster pace than gross profit, etc. I can determine none of those accurately unless I can get stock option comp expense in the income statement. What if Company B suddenly switched from using cash compensation to restricted stock? Without options being expensed, Company B would magically show increased operating income, increased operating margins, lower SG&A expense, etc. It would like an amazing quarter, when in reality all it did was change the method in which it compensated its employees. That doesn’t make sense to me.

And options are an expense. No, they do not cost the company any immediate cash. But they do cost shareholders via diluting per share stockholder equity the same way a cash payment would reduce shareholder equity through lower net income for the period. And every share or option granted to an employee is potentially a share that could have been sold in the market resulting in cash to the company.

To use BOFI as an example, every quarter management sells a certain amount of stock “at the money” into the market. The sales raise cash that the company is able to put to use to grow the business. If the company were to grant those same shares to management, instead of selling those shares to the market, the company would receive no cash yet indure the same share count dilution. Existing shareholders would get doubly hosed with a reduced per share stockholders equity AND no additional cash for management to use to grow the business. So there is a very real cost to share and option grants, and I want to see that cost in the company’s operating results.

OK, end of rant, and off my soapbox. Back to the BOFI conference call.

Fletch

13 Likes
8

Hi again Fletch, I was writing to Andy there, and didn’t mean to upset you. Let’s just agree to disagree.

Saul

9

Thanks Saul, I appreciate it and that makes it clearer. I guess I just wasn’t questioning Gaap and that’s why I couldn’t see the answer.

Andy

1 Like
10

Thanks Denny sort of like double jeopardy eh? (pronounced aye)

Andy

11

Fletch,
Thanks for your answer. This is how we all learn when people discuss different view points. We all need more of these discussions.

Andy

12

Excellent discussion!

Anirban

1 Like
13

Why, oh why, am I weighing in on this? I do not really care . . . .

Scenario A. Company A pays its CEO $1,000,000 in cash as salary. He turns around and invests that cash in the company, buying newly issued shares.

The company ends up with the same $1,000,000 cash it started with and now has $1,000,000 worth of extra shares outstanding and a happy CEO. I assume there is no question at all that here the earnings should be reduced by $1,000,000 and the share count should reflect dilution by $1,000,000 worth of shares. I.e., the dreaded double hit to EPS occurs. And quite appropriately here, right? (The books should also reflect additional capital of $1,000,000.) I doubt that anyone would quarrel with this accounting treatment.

So why should things be different if we just collapse the salary and share purchase into one transaction?

Scenario B. Same as Scenario A, except that instead of receiving cash salary of $1 Million and then reinvesting it in company newly issued stock, the CEO just says, “Let’s save ourselves a step – just hold onto the cash and give me the shares directly.” Everything ends up exactly the same as above; but Saul would now want to avoid the dilution of EPS or the hit to earnings.

Why should these two scenarios – both starting and ending in the same place – be accounted for differently? That would be elevating form over substance, right? Or am I incorrect in my above discussion (very possible).

Rich

CED

Who generally ignores numbers and just invests by “feel” . . . .

And who, perhaps not coincidentally, has the approximate inverse of Saul’s long term stock performance . . . .

9 Likes
14

Sorry Chance, My last post , which is on a different thread, is an answer to you.

Saul

15

Rich,
Why in scenario A would the company have to issue more shares? The CEO goes out on the open market and buys shares on their own and then no new shares would be issued.

I think that Fletch makes a valid point but so does Saul. Why not follow Gaap till you get to EPS an then use Non-Gaap? That way the cost of the shares would be counted for your margins but then when it comes to EPS you are not counting them twice.

I am not an accountant just trying to think what might be a better way to track my companies to more accurately reflect how their business is doing.

Andy

16

Why in scenario A would the company have to issue more shares? The CEO goes out on the open market and buys shares on their own and then no new shares would be issued.

Hi Andy,

The company does not have to sell shares to the CEO; I am just saying, suppose it does. If it does (and such things happen from time to time), then sure it keeps the cash, but as a capital contribution, not as earnings. And I think it is clear this is how it would be treated for financial accounting purposes (I know that is how it would be treated for tax accounting purposes – I was a tax attorney for many years).

Would it make sense that if Scenario A did happen (for whatever reason), there should not be a hit to earnings for the salary paid to the CEO? I think people would probably agree that there should be a hit to earnings in Scenario A, right, if for some reason that Scenario did in fact occur? (We have to keep the idea of cash flow separate from the idea of P&L, remember.)

So, if we agree that Scenario A should result in a hit to earnings (in the unlikely but possible event that Scenario A were to occur, why treat things any differently if the CEO just gets shares as compensation directly (Scenario B)? Everything ends up exactly the same in Scenario B as in our hypothetical Scenario A, right? So it seems (to me) that there should be a hit to earnings and a capital contribution for shares in Scenario B as well.

Of course the cash situation in both Scenario A and Scenario B are in the end the same as if the company just kept its $1 Million profit, but that is why one looks at the cash flow statement, for that kind of info. Capital contributions should not be treated as earnings just because they are both cash. The company is paying the CEO a salary of $1 Million (in stock, but it is still compensation) – that is a very different P&L proposition than if the CEO is working for free. We need to be able to distinguish those two situations when we read the financial statement (in the very unlikely event that I ever read a financial statement).

To make it even stranger, suppose (Scenario C) that the company pays a $1 Million dollar salary to its CEO and then sells a million dollars worth of shares to YOU or, say, Tom Brady. The company is in exactly the same cash position as it is in Scenario A or Scenario B (where it gives shares as compensation to its CEO) – but of course no one would say there should not be a hit to earnings in Scenario C, right?

So why let form dictate the results over substance? The substance in all of these cases is compensation and a capital infusion; I fear that it is misleading to allow the company to act as if it is earning its profits without paying its CEO just because it pays him with stock instead of cash. This would allow the company to hide its CEO’s compensation by, in essence, treating a capital infusion as a profit, basically. But that is JMO, and I am a nobody, so honestly, who cares what I think?!

Also, to be clear, I am just discussing this stuff because it is late on a Friday night, my kids are all away in distant lands doing fun things, and it is honestly quite lonely here. But I never thought I would reach the point of discussing accounting issues on a Friday night because it beat all of the available alternatives. Wow!

Anyway, please excuse any conclusory or emphatic nature of the above. I really know very little in this area, and I also bring zero emotion to this issue. It is just really hard to be nuanced in a board post!

Thanks for all your fine work on this and other boards – and of course thanks to Saul, who has added an immense amount of value to the Motley Fool universe.

Rich

CED

10 Likes
17

Hi Rich Chance CED, Lying in bed I figured out the answer to your puzzle and I just had to get up and write it down before I forgot it.
Here it is:

Scenario A. Company A pays its CEO $1,000,000 in cash as salary. He turns around and invests that cash in the company, buying newly issued shares. The company ends up with the same $1,000,000 cash it started with and now has $1,000,000 worth of extra shares outstanding and a happy CEO. I assume there is no question at all that here the earnings should be reduced by $1,000,000 and the share count should reflect dilution by $1,000,000 worth of shares. I.e., the dreaded double hit to EPS occurs. And quite appropriately here, right? (The books should also reflect additional capital of $1,000,000.) I doubt that anyone would quarrel with this accounting treatment. So why should things be different if we just collapse the salary and share purchase into one transaction?

Okay, the company pays the CEO $1 million in CASH. At this point they have $1 million less earnings. Period! Full Stop!!! (It’s less earnings because they paid him cash, not stock, which is non-cash. They no longer have the money).

Then they decide to sell some shares in a secondary. Whether they sell no shares, a half million dollars worth, or $1 million, $2 million or $3 million dollars worth doesn’t matter. It’s still a secondary. Whether the CEO buys the shares, or someone else entirely, or a hedge fund, doesn’t matter, it’s still a secondary. What confused us was that you happened to set the amount that they were selling as the same amount that they paid the CEO, but conceptually the two transactions ARE NOT related.

Thanks for the interesting puzzle, but no thanks for waking me up in the middle of the night!

:wink:

Saul

2 Likes
18

In a similar discussion yesterday I posted:

financial analysis is not one of my strengths. I’m pretty good at looking at a business operationally, but not financially.

It’s not the arithmetic, I can do that with my eyes closed, both hands tied behind my back and no calculator - it’s the damn financial statements. GAAP is intended to provide a consistent way of comparing apples to apples across all different businesses. This might be an inherently flawed concept to begin with.

This thread pretty much highlights most of my problems with financial analysis - it’s just damnably difficult to know what’s important. So, I tend to focus on the company’s business. As I said before, I don’t ignore the financials completely, I just lean heavily on people like Saul and Fletch and others who know enough about it to have an intelligent disagreement about the methodology.

What I would find really interesting is to know how often you disagree on the conclusion regarding the future prospects of an investment. What I mean is if you to disagree about GAAP v non-GAAP but in the end pretty frequently reach similar conclusions, then it’s rather a moot argument.

2 Likes
19

I wrote the above in the middle of the night. Here’s a slightly more coherent version:

Okay, the company pays the CEO $1 million in CASH. At this point they have $1 million less earnings. Period! Full Stop!!! (It’s less earnings because they paid him cash, rather than stock which would have been non-cash. Since they paid him cash, the company no longer has the money. It’s a real expense.)

Then they decide to sell some shares in a secondary. Whether they sell no shares, a half million dollars worth, or $1 million, $2 million or $3 million dollars worth doesn’t matter. It’s still a secondary. It’s irrelevant whether the CEO buys the shares, or someone else buys them, or a hedge fund buys them. It’s still a secondary and doesn’t affect earnings. What confused us was that you happened to set the amount that they were selling as the same amount that they paid the CEO, but conceptually the two transactions ARE NOT related. If you said the company decided to sell $2 million in shares we might not have been so confused.

Hope this makes it a little clearer.

Saul

1 Like
20

What I would find really interesting is to know how often you disagree on the conclusion regarding the future prospects of an investment. What I mean is if you to disagree about GAAP v non-GAAP but in the end pretty frequently reach similar conclusions, then it’s rather a moot argument.

Good point BrittleRock! Fletch and I actually usually agree on the stocks (as far as I know), so I guess it IS a moot point.

Saul

1 Like