Quote from optionsgirl:
I gather from my lay experience that there are less surprises in earnings release. However, it is hard to come by with any statistical proof. I would think this is an old idea where one wants to construct a strategy if there was a significant bias towards earnings estimates being right or wrong when quarterly earnings are released... Perhaps it depends on the company itself? I would think some companies are more predictable than others. Anyone know of any book or site that elaborates about all this?
I believe you'll have great difficulty constructing a profitable strategy based on the probability of a company beating or missing earnings.
The problem with this is that an earnings beat or miss does not predict the resulting price action. There are many factors that weigh on the market's reaction to an earnings report, mainly gleaned from the earnings conference call such as future earnings and revenue guidance, operating margins, the effect of current economic conditions on company strategies, etc.
Then you also have to take into account the price action leading up to earnings. Are market expectations priced in already, meaning even an earnings beat may result in "buy the rumor, sell the news" price deflation? RIMM for example just beat earnings, but they'd run up so much prior to earnings and expectations were so high that anything less than stellar in the conference call resulted in a large selloff.
I once bought a put on company because I saw 2 million shares sell off in 2 minutes the day before earnings so I assumed someone knew something. Plus price had run up a lot so I figured there would be a "sell the news" correction. They reported a triple play - earnings beat, revenue beat and upside guidance and price opened gapped up $4 a share from the previous day's close. Price pulled back a bit on the open and I sold my put at a very small loss, figuring it would continue to lose money as it was out of the money near expiration. Price then continued to fall non-stop the entire day, dropping $13 a share before finally finding support. Had I waited I would've had a $200 gain.
I ended up making $670 on the bounce off support late that day, proving that the safest way to a profit is to trade the price action that follows earnings report after the market opens.