Ok, so it is sort of cheating to post work on which I am a co-author, but given I just spent the entire afternoon on the phone, I am short on time, so this will have to suffice.
SSRN-Forecasting Power of Implied Volatility: Evidence from Individual Equities by Jonathan Godbey, James Mahar:
1) implied volatility is a better forecaster of realized volatility than either historic volatility or GARCH models and 2) the information content of implied volatility significantly decreases with liquidity. "
The paper is interesting for several reasons. First and probably most importantly, we show that IV is still the best forecaster of future realized volatility. Secondly we find that the information content of implied volatilty does drop with volume. Previously this had been shown before for index options and for a small collection of firms, but to the best of our knowledge, not for a sample as large as ours. Finally, we show that the IVs derived from call prices is almost identical in information contect as that taken from put prices.
Godbey, Jonathan M. and Mahar, James W., "Forecasting Power of Implied Volatility: Evidence from Individual Equities" (July 14m 2005). http://ssrn.com/abstract=762644
BTW I was going to wait until it had been "reviewed" but figured I would post it now.