I'm seeing more light being shone on the trend of analysts quietly lowering eps expectations prior to earnings, to set up easy "consensus beats" and good headlines. As the chart in here shows, the average eps cut prior to reporting in the last 4 years has been 4%, and the average consensus beat has been 3.3%. You can imagine the impact to price multiples if all those earnings had missed expectations instead of being "better than expected". I doubt the practice will change though, since there's clearly a symbiotic relationship between analysts private access to company info and steak dinners, and their public love for the company.

zerohedge.com/ne...-chart