Beware of GAAP Again

Includes: DIA, SPY
by: Marc Gerstein

When times are tough investors and members of the financial media love to bash analysts and corporate executives, especially when it comes to issues relating to accounting. After all, it's still less than a decade since some once-mighty corporations were brought low by accounting scandals. But please, please, please don't go overboard and assume that accountants are the ones who are always providing the most helpful information. Notwithstanding some occasional abuses (every profession has some bad apples), security analysis in general is honest work and can provide insights that are much more useful to investors. Unfortunately, however, it's GAAP that sometimes obscures economic reality.

Back in June 2009, I wrote about the problems with the way operating profit is reported, specifically, how this item can be heavily distorted by unusual gains and expenses. I showed why it's important for analysts to create an alternative measure of profits that provides a more useful basis for making assumptions about the future, a practice inspired directly from the most revered masters of security analysis: Graham and Dodd.

Today, I want to explore a different aspect of this issue, the impact on trailing 12 month (TTM) EPS, this being a figure widely used by pundits who discuss trends in corporate earnings in general, and earnings-based stock valuation in particular. This is not to say that everybody with a soapbox talks about TTM earnings. Knowledgeable investors and commentators prefer forward-looking estimates which do not suffer from the same informational defects. Often, though, those who most loudly trumpet TTM earnings tend to have agendas, usually involving a desire to beat up on analysts, to beat up on stock valuations, or both. It's important that investors look critically at such commentary, much the way they needed to do so back in 2002-03, the last time Wall Street critics bombarded us with rhetoric about how crazy it was for stock prices to climb when earnings were so horrible and about how analyst projections for '03 and '04 were so far out of line.

As I wrote in June, I do not dislike accountants. The issue I raise reflects the nature of the accounting missions; tracking and disclosure. These are critical. We need to know what the company has done and have confidence in the truth of the facts presented to us. But for investors, this is only step one. Ultimately, we live or die based on the assumptions we make about the future. The type of information we need to do that, and the nature of the presentation we require, can differ quite substantially from that which is necessary to track and disclose.

EPS trends - a first glance

Looking at analyst estimates for the current fiscal year, it appears that trends have stabilized following the brutal series of downward revisions we saw in 2008.

Figure 1

Next, look at Figure 2. It shows how the estimated results compare to TTM EPS.

Figure 2

The data sets are not strictly comparable. TTM measures the last four quarters which at this time, for calendar-year companies, would likely be the 12-month period ending 6/30/09 or 9/30/09 depending on who reported the third quarter and who didn't yet do that. Meanwhile, the estimate is for a fiscal year; for calendar-year companies, it's the 12-month period ending 12/31/09. But the disparity doesn't cloud the ultimate point, to wit, that analysts are, on balance, projecting massive gains above the TTM levels. Numerically, we're talking about a projected gain of 638% for the S&P 500.

Are these expectations plausible?

That's a loaded question. Actually, nobody really believes that economically speaking, S&P 500 companies are going to see earnings gains anywhere near that magnitude. Analysts and professional investors are looking at much more moderate percentage gains with differences among them based on how each chooses re-cast the historical figures.

Earnings as an art as well as a science

The question for today is whether it's reasonable to making any sort of adjustments to reported earnings. A decade ago, this practice became much reviled as legitimate accounting phrases like pro forma earnings gave way to justifiable sarcastic variations such as EBE (Earnings Before Everything or Earnings Before Expenses).

Forget the old labels. Pro forma is a technique that can be used to project backwards to see how two separate companies would have performed had they been merged into a single entity. It's most common use is to help investors assess proposed corporate mergers. The phrase was used inappropriately back in the day and is completely irrelevant to today's topic.

To get to the heart of the issue, start with Table 1, which summarizes the magnitude of the TTM EPS percent changes among companies that experienced declines.

Table 1

Those are incredibly brutal numbers and seem to mesh well with the red trend line we saw in Figure 2.

So what, exactly, is going on here? Was business really that bad? Let's look at Table 2, which shows the same information for Revenues.

Table 2

What a difference! The magnitude of the declines here are certainly noteworthy, enough so to make it obvious that the economy has been troubled. But the revenue changes are nowhere near as dramatic as the nightmare we saw with EPS. So let's go down the income statement a bit and see if we can locate the land mines were.

Table 3 shows the trends in gross profit which, generally speaking, is revenue minus cost of sales; the costs most directly associated with the production of revenue (i.e. raw material, salaries of employees involved in producing goods or providing the services, etc.).

Table 3

Clearly, we see some deterioration. And that's to be expected. Many items classified under cost of sales are variable, but in life, the variation with revenue is not strictly one-to-one. The decline in these costs is usually less than the decline in revenue, hence the greater shrinkage in gross profit. So to this point, we're following the script.

Let's look at Table 4, which shows operating profit. This is gross profit minus selling, general and administrative cost (often referred to as overhead) and depreciation. These items are more fixed than is the case with cost of sales. But the magnitude is not usually that great, so the change in operating profit would, one would think, move more or less based on changes in gross profit (with a bit more of a decline in operating profit).

Table 4

Ouch! That's disaster is grossly disproportionate to what we'd expect based on what we saw with gross profit. But it is very much in line with the mess we saw with in Table 1 for EPS.

It looks like we found the land mines. Something happened between gross profit and operating profit. And whatever it was, it was way too large to be attributed to traditional operating expense items (selling, general and administrative expense and depreciation).

The answer: "unusuals." One example would be losses on the disposal of a business. Another might be expenses incurred in connection with the shutdown of a subsidiary or plant. It might be an accounting entry to adjust for the diminished value of an asset. This is not an exhaustive list of possibilities, and unusual gains happen too. But you get the idea. The common element is the unusual nature of these items.

Who, if anyone, should we criticize?

For analysts, recalculating earnings without these unusuals is second nature. Graham and Dodd told them to do it. But it wasn't a hard sell. Common sense dictates it as well. Analysts are looking for those elements of the financials that help them formulate expectations for the future. Unusuals don't fit. It's important that accountants calculate and disclose them. But when we're formulating future-oriented investment expectations, we need to put them on the shelf. (See prior article for more on this topic.) So don't criticize analysts for doing it.

Similarly, it would not be appropriate to criticize management for logging these unusuals. Life is complex. When something odd happens in a business, whether its due to management mistakes, events beyond management's control, or whatever, we want it properly recorded.

If there's to be a beef with anyone, perhaps it's the standard-setting body within the accounting profession.

There is a category of items referred to as extraordinary. Those who design financial databases all understand this and are quite proficient at presenting Earnings and EPS excluding extraordinary items. The problem is that the extraordinary category is defined very narrowly, meaning it's very hard for management to classify an item under this heading (the cumulative impact of a change in accounting policy would be one example). If we're to get the red line in figure 2 to make analytic sense, it would probably require a formal re-definition of extraordinary items.

Interestingly, a February 2007 article in The CPA Journal proposes the opposite solution: elimination of the Extraordinary classification. Even more interestingly the author cited the same problem I referred to; the excessive narrowness of the category and expressed particular frustration about a ruling that went so far as to prohibit companies from reporting losses relating to the September 11, 2001 attacks as extraordinary. I definitely share the author's feeling about the issue, but I differ with his conclusion. to an accountant, the Extraordinary category is little more than a seldom used irritant. But to an investor, it's a vital but horribly under-utilized classification that needs to be properly used as a way to categorize exactly what the label suggests it ought to categorize; items that are not ordinary. And in my thesaurus, ordinary is a synonym for usual. Therefore, extraordinary means unusual.

So let's stop bashing analysts and corporate executives and start advocating for change where it's needed, with Generally Accepted Accounting Principles.

Disclosure: No positions