Sorry for bumping an old thread, but I stumbled across this old thread that I posted on a while back, when I was searching for something else. I just wanted to set the record straight here. I was completely wrong and dtrader was right. I talked to a statistician a while back about this and he was able to explain why I was wrong in several different ways that I was able to understand. (I don't remember exactly well enough to explain in my own words though, so this is clearly an issue I still have some difficulty understanding.)
No, you were right and dtrader was wrong. Your statistician friend is mistaken. For a fuller explanation see the father of moder finance, Eugene Fama: http://www.dimensional.com/famafrench/2010/04/qa-semi-variance-a-better-risk-measure.html