Does the option market produce superior forecasts of noise‐corrected volatility measures?