Forget "Analyst Estimates"
1) Estimates about future earnings are almost always far too optimistic.
2) Estimates about earnings in the quarter that just ended are almost always far too pessimistic (the only exception that I can recall were for financial companies in 2008 due to large write downs).
Why do these systematic errors arise? Well, simply because these "analysts" are employed by companies that want people to buy stocks. It suits their purposes if people hear that "earnings are better than expected" and at the same time hear that rapid earnings growth is projected in the future. No matter how bad the earnings really were in an absolute sense, hearing that they were "better than expected" and will increase in the future will make people unaware of the phony and biased nature of the "estimates" more willing to buy stocks. This is the purpose of these so-called “estimates”.
The current earnings season have proven to be no exception. Just as I predicted before it started, most companies have reported falling earnings, but since earnings were higher than "analyst estimates" in most cases, it was still interpreted by many as a reason to buy stocks.
Of course, the move from 1) to 2) requires that estimates are revised down significantly in the meantime, but that is made in a very quiet way that few notices.