Apple's 92% earnings windfall: The bloggers nail it, the pros miss by a mile

The shortfall in iPad sales was one of the few categories the Street called correctly

The Wall Street analysts who get paid to cover Apple (AAPL) got some things right.

Several predicted Apple's total revenue for the second fiscal quarter of 2011 within a few hundred million dollars. And their tendency to underestimate Apple's unit sales served them well in the iPad category.

But when it came to the number that matters most -- the bottom line: earnings per share -- it was the amateurs, once again, who nailed it and the pros who blew it.

In our ranking of the best and worst analysts for Q2 2011, which lists them by how accurately they predicted Apple's revenue and EPS, the amateurs took 13 of the top 15 spots.

The bottom 32 spots were all held by professionals working for banks and brokerage houses. Taken as a whole, the numbers they sent their paying clients were off by a margin (11%) more than four times as big as those generated by the guys who do it for free (2.5%).

The bloggers weren't quite as prescient as they have been in past quarters. None of them foresaw that what COO Tim Cook called "the mother of all backlogs" would hurt iPad unit sales so badly. And as a group they overestimated Apple's total revenue for the quarter. It was a Wall Street professional, Caris' Robert Cihra, whose revenue estimate was closest to the mark.

But as a group, the professionals have nothing to be proud of, as the distribution of green (for good) and red (for bad) squares in the chart below the fold shows at a glance.

Green: best estimate. Red: worst. Light green: 2nd or 3rd best. Pink: 2nd or 3rd worst.

Mean absolute % error by category. Source: Daniel Tello

Candidates for the analyst's hall of shame:

• Goldman Sachs' Bill Shope, who underestimated Apple's revenue by nearly $2.7 billion and iPhone sales by 5.4 million units and came in dead last in the rankings.
• Hudson Square's Daniel Ernst, who underestimated Apple's EPS by 22% and its gross margin by 301 basis points.
• Gabelli's Hendi Susanto, who overestimated iPod sales by 2.3 million units and iPad sales by 4.1 million units and tied Ernst for the worst gross margin prediction.
• JMP's Alex Gauna, who made a name for himself last month by downgrading Apple in the middle of the company's best non-holiday quarter ever. He fell off our charts because he never supplied the unit sales numbers we requested, but his revenue estimate of $22 billion would have been the second worst, and his $5.1 EPS estimate not much better.

Special mention on the amateur side goes to:

• Financial Alchemist's Turley Muller, a veteran Apple watcher who nailed the Mac unit sales and gross margin numbers.
• Asymco's Horace Dediu, another veteran, who tied J.P. Morgan's Mark Moskowitz for the best iPhone estimate
• Luke Kittell, a newcomer from the The Mac Observers Apple Finance Board (AFB), who nailed the earnings per share number.
• Two more AFB members, Jeff Fosberg and Adam Thompson, who grabbed the No. 1 spots in the best Rev & EPS and All Categories lists, respectively.

Special thanks to Robert Paul Leitao of Posts At Eventide for curating the AFB submissions and to Daniel Tello of Deagol's AAPL Model, who corrected my math errors and provide the spreadsheet above showing how the gap between the pros and the amateurs is widening -- at least in most categories.

"Friggin' iPads keep making us look bad," he says. "Oh well."

Also on Fortune.com:

• Apple clobbers estimates, iPad sales fall short
• Apple earnings: 5 numbers to watch
• Apple Q2 2011 earnings preview

[Follow Philip Elmer-DeWitt on Twitter @philiped]