By definition, historical volatility will always be backward looking and lag the real-time volatility environment. When volatility trends, as it has over the course of the past two months, this phenomenon is less evident. In the current market environment, however, where both stocks and implied volatility measures such as the VIX are reversing, many measures of historical volatility begin to seem no more useful than driving on a winding road using the rear view mirror.
The graph below shows that while the 30 day historical volatility in the SPX peaked a week ago, it has only fallen 4% since that time. Not surprisingly, the 50 day historical volatility measure has been even less responsive to recent market changes and has entered a plateau phase, seemingly oblivious to the recent changes in volatility. By contrast, the 10 day historical volatility measure has dropped 27%; it has more accurately mirrored the VIX and perhaps even hinted at a coming VIX reversal when it topped three days before the VIX did.
Ten day historical volatility is erratic and prone to readings that reflect a low signal to noise ratio. When volatility is changing quickly, however, 10 day historical volatility is frequently a better proxy for the VIX and a more meaningful baseline for comparison than its 30 day sibling.

by Bill Luby (VIX and More )