Those focusing on EPS numbers, take note of the following article in Money Daily.
New earnings-per-share reporting rules are a change for the better
The problem: Right now, they're creating a big mess
by Michael Brush
Harry Truman once quipped that he only wanted to work with one-armed economists. That way, he reasoned, they couldn't fudge things by saying: "On the other hand..."
Investors are looking for the same thing in accountants this earnings season. Like Truman's economists, company accountants and Wall Street analysts are coming up with two different versions of that
all-important measure of company performance -- earnings per share.
The situation -- caused by a new change in accounting rules -- is causing a lot of confusion over whether companies are meeting or
beating expectations. As a result, companies can get punished when they actually did pretty well, and vice versa.
Before we go into the details behind this mess, it is important to keep three things in mind. First, this will be the last quarter that you have to deal with the problem - by next quarter, analysts and
companies should all be on the same page. Second, if you are judging companies on the basis of how they measure up to expectations, be very careful: The numbers this quarter may not be what they appear to be.
Third, the change in accounting rules is actually a good thing. It means that from now on, when U.S. companies calculate earnings per share, they will be taking into account things like options,
warrants and convertible bonds. These types of securities can be
converted into stock ... and lower a company's earnings per share.
"The new 'diluted' numbers give investors the best look they have ever had at the real earnings per share," says Ed Keon, of IBES, a firm that reports on earnings estimates. "It includes all of what is out there in terms of who may convert options."
Still, it's a new system, and it's causing a good deal of confusion this quarter. Under the old rules, which changed last December 15, companies could often ignore instruments that could be converted into shares when calculating earnings per share. The rules said that if considering all the options, warrants and convertible bonds to be actual shares did not dilute earnings per share by more than 3%, then you could just pretend those things did not exist.
The old system gave companies three choices in deciding how many "shares" were out there when they calculated earnings per share. They could report simple earnings, or net income divided by the number of common shares outstanding. They could also use primary earnings, which included most options -- in addition to common shares -- in the calculation. Or they could use fully diluted earnings, which included all options, plus things like warrants and convertible debt.
Starting with last quarter's earnings, however, companies now have to report two kinds of earnings. They are basic earnings (which is similar to simple earnings under the old rules) and diluted earnings (similar to fully diluted earnings under the old system).
This caused confusion this week over IBM's earnings. Most investors, not to mention earnings tracking services like First Call and IBES, thought Wall Street analysts were using diluted earnings when those analysts came up with their consensus estimate of $2.15 a share for IBM, for the fourth quarter of last year. So when IBM reported $2.11 diluted earnings and $2.16 basic earnings (a higher number because it excluded options), it looked like the company missed consensus estimates. It turns out, though, that the analysts were really reporting basic earnings when they made estimates. So IBM really beat the consensus by one cent, or $2.16 to $2.15. IBM shares traded down after the announcement, in part because of this confusion.
What's worse, the snafu is likely to happen again during this earnings season. That's because no one has really gone through to
check all the analysts' estimates to see if they are based on diluted
or basic earnings. The general agreement between Wall Street
analysts and firms like First Call and IBES is that the analysts
should be reporting diluted earnings in their estimates. But as the
IBM example shows, this is not always the case. Here's why the chances are good that there are more surprises out there. Given that there are about 3,000 analysts reporting on about 6,000 companies, it would simply be too time consuming for the services like First Call and IBES to find other exceptions, especially in the midst of busy earnings reporting season, when analysts don't have much free time to explain things to them. "The brokerage firms told us the numbers were diluted," says Keon. "But it turned out that for IBM they were basic. For the fourth quarter I would not rule out further misunderstandings."
First Call's Chuck Hill agrees the problem could come up again. "For the fourth quarter there could be more basic estimates in there, and there could be more IBM-type surprises. The bottom line is that it is fuzzy this quarter." And even when it is clear that there is an apples to oranges comparison, some money managers may not pick that up -- and sell off a stock that looked like it missed earnings when it really did not. "I am afraid that there may be someone out there who has some clerk pulling numbers off some feed and just shoving them into a computer model," says Hill. "And if that is what you are doing, you have some problems." Here is another problem investors will run into because of the change in the accounting rules. Companies have gone back and restated their 1996 earnings under the new rules, but the numbers for years before that may or may not be restated. So if you judge stocks by looking at things like five-year earnings growth rate, you should keep that in mind. The good news, however, is that the difference probably won't be that big. And it can't be bigger than 3% a year. After all, if the dilutive effect of options and convertible securities was greater than 3% in past years, the companies would have been reporting the watered-down numbers all along...
Ram |
