I use seasonally adjusted regression of past revenues to forecast future results. Estimate net profit based on profit margins and EPS from there. I then estimate a fair price using Ben Graham's value formula.
First, BIDU's revenues yesterday were 2.2% below the predictable historic trend. Net profit and EPS were above what this method predicted. Anybody that was surprised wasn't looking at the data. The quarter reported has been the lowest profit margin quarter of the year for the last 16 quarters (averaging 23% versus 47%, 40% and 30% for the next three, if history repeats itself.)
Next quarter, expect:
Revenues - $3.2B
Net Profit - $680M
EPS - $1.95
Estimated fair value - $220 using a growth rate of 12%, the rate at which net income has grown over the last three years. The current price indicates a future growth rate of 7.7%
The R^2 on the regression is 0.993
This is the most conservative of the predictions I can make from my regression of the raw data. Revenue has been growing more than 40% per year. Net profit margins for the next quarter are more likely to be 40% than 23%. These assumptions would put the fair price 2-3x higher. The market has over-reacted, which significantly lowered the risk of buying the stock now.
I welcome critique of my method or new or corrected historical data.