In this particular case (option prices and market bias) there is evidence that the slope of the skew in SNO has some value in predicting the underlying stock returns. There are a few common sense explanations to that. PS. I dont't think the skew has any real predictive value in the index space