[ddoonie]

I must agree with this method. There is no substitute for researching companies. I'd filter a bunch of 10k's and look for abnormalities etc.. You can start with the typical top down approach - decide what your view is on the economy, pick an industry you think is poised to do well or badly in those circumstances, start filtering firms that fit in with your assumptions. Look for the drivers of revenues and costs for these firms and see if you can find anything that is 'abnormal' and make a play based on those...