Thanks for the answer. The method is a rule of thumb, I guess, but surely no exact science. When you consider that I compare percentages and not real values, it is not twice but even three times removed from reality;-)

What gives me confidence in the number is the fact that we both came to the same conclusion using very different methods (you: the price and it's peaks, I: the average volatility). This gives the whole assessment a robust impression, as I think.