Wednesday, July 29, 2015

McClatchy earnings shows limits of automated stories

There's been a lot of ballyhoo about AP's use of computer algorithms to generate hundreds of earnings stories.

Among AP's reasons was that it could provide much wider coverage. Reporters would still handle the major stuff, the wire service said.

At some point, however, the question of quality vs. quantity was going to raise its head. And here's an example of where the automated system fall short. Here's the AP's auto-generated story on McClatchy's recent earnings.

Pretty bare bones stuff. But this isn't a plain-vanilla situation. In fact, there's some serious insight here. This is one of the old-line pure-play media companies and in many ways is a barometer of how midmarket newspapers are likely to fare. And there are, after all, about 62 million shares outstanding, with Yahoo Finance saying that as of the end of March, 119 institutions held shares. That means more than a few people have these shares in their retirement and other accounts (and may not realize it).

Here's another version that, I think, is more reporter generated:

Those second, third and fourth grafs contain some important context. It's not just that the company eked out a profit. It's that the stock's price has plunged about 60 percent since February as it became apparent those earnings -- any earnings -- were generated largely through throwing the ballast overboard on a very leaky ship. So if you read the AP story, you come away with "they made money -- a small bit, but still a profit." Read the other one from American City Business Journals and you'd come away with more understanding and, perhaps, many more questions.

There are, I think, going to be a lot of these kinds of stories in the midrange of companies not really big or sexy enough to draw the AP's resources, yet large or important enough in their own way that they deserve more contextual treatment. So, even more so, investor beware and understand the limitations of what AP is doing

Labels: , , , , ,

0 Comments:

Post a Comment

<< Home