Financial websites publish crappy statistics

I use websites like google and yahoo to screen stocks, but I think it is a bad idea to put too much faith in the numbers they report. Willi Foods has a P/E of 6.6 according to Yahoo!, but this includes one time gains in the past year that will not be repeated in the coming year. But this is fairly harmless in comparison to their practice of not deducting stock option expenses. Infomation Arbitrage cites writer Herb Greenberg in this article.

The trouble comes when analysts, for whatever reason, don't include options expenses in their primary earnings estimate. Their numbers are then compiled -- and consensus is created -- by three companies: Thomson Corp.'s Thomson Financial, Reuters Group PLC's Reuters Estimates and Zacks Investment Research. Zacks, which claims to have created the industry nearly three decades ago, is the only one whose consensus always includes options expenses, even if it means manually figuring it out; as of the third quarter, it had to do that for 300 of the 4,500 companies in its database. Reuters Estimates and Thomson Financial base their consensus on majority rule. "We need to conform to market standard," says Ashwani Kaul, senior research analyst for Reuters Estimates.

********************

And that's where this story gets weird. Only a handful of brokerage firms, including Bear Stearns; Bank of America Corp.; Goldman Sachs Group Inc.; and Sanford C. Bernstein, a unit of AllianceBernstein LP, require analysts to include options expenses in estimates that are picked up by Thomson Financial, Reuters and Zacks.

Why don't all of them? Good question -- and viewing earnings any other way is puzzling even to the CFA Institute, which charters analysts. "Our view is that options are a form of compensation," says Rebecca McEnally, a spokeswoman for the institute's Centre for Financial Market Integrity. "And the fair value of options expense should be treated as any expense in the forecast. To fail to do so is to underreport expenses and overreport the earnings." If that's the case, then why are we having this discussion?

There is no silver bullet to the problem of unreliable statistics. However, I'd like the financial analysts to present the information as accurately as possible. I'm considering that I ought to change my data providers to those that treat options as an expense.