Anybody who follows its earnings reports can tell you that Apple (AAPL) guides conservatively, under-estimating its forward-looking numbers to demolish them at the end of the quarter.

Only once in the past 23 quarters — in the summer of 2006 — did it miss its revenue guidance, by 0.7%. In the other 22 the company beat its forecast by anywhere from 1.2% to 35.2%. Last week’s beat: 19%. Average beat: 10.9%. (See chart.)

So the big surprise in Apple’s third quarter earnings report Tuesday was not that it guided conservatively, but that its revenue forecast was so monstrous. Last year Apple took in $12.2 billion in its fiscal fourth quarter, which ends in September. This year the company told analysts to expect Q4 revenues of $18 billion — a 47% increase.

“We hadn’t girded ourselves for a mammoth revenue forecast,” wrote Oppenheimer’s Yair Reiner the next day, “and neither, we believe, had the Street.”

Like every other analyst who follows Apple, Reiner immediately hiked his Q4 revenue estimate. But he didn’t forecast anything near $19.96 billion, which is what $18 billion plus Apple’s average beat comes to. Instead, Reiner raised his estimate to $18.5 billion from $16.6 billion. The Street’s revenue consensus for Q4, according to Thomson Financial, is $18.29 billion — a measly 1.6% beat. Apple has underperformed at that level only three times in the past 23 quarters.

In other words, Wall Street is either betting that Apple was hyping its revenue forecast, perhaps to send a message about the death grip flap. (“Antennagate schmantennate,” is how Reiner put it.) Or the analysts, in a time-honored sell-side ploy, were doing some low-balling of their own so clients who follow their advice and buy Apple shares might be pleasantly surprised come the end of the quarter.

Below, the estimated percentage beats of the 15 analysts we polled, starting with Gene Munster’s timid 0.0%. The average among this group is 2.4%. The outlier is RBC Capital’s Mike Abramsky, at 9.5%.

FYI, Reiner came in 3rd in our quarterly rankings of the best and worst Apple analysts last week. Abramsky was 12th, and Munster 21st out of 32.

ADDENDUM: Financial Alchemist‘s Turley Muller, a blogger-analyst whose Apple estimates regularly beat the professionals’, writes that if you adjust for last December’s accounting change, Apple’s Q1 2010 guidance beat comes out to 18.3%. That’s quite a comedown from the 35.2% we’ve got in there now. Even so, if the 19.3% beat he figures Apple has averaged over the past three quarters were to continue, the company might report revenues of nearly $21.5 billion in Q4. “A number like that is too mind-blowing to even think about,” says Muller. His current estimate: $19.5 billion, which might go up as the quarter progresses if Apple’s iPad and iPhone 4 supply issues resolve themselves in time.

See also:

• Apple’s earnings: Antennagate Schmatennagate
• Earnings Smackdown: The best and worst Apple analysts
• Apple Q3 blow-out: Profits up nearly 78%

[Follow Philip Elmer-DeWitt on Twitter @philiped]