If realized is predictably lower than implied then options are predictably overpriced and you could obtain abnormal returns selling them. The literature certainly doesn't agree with your assertion except in a couple well defined areas (otm s&p 500 puts are one example) and its something thats easy to test empirically, so I'm curious what your data source on that assertion is? BTW, if not options what else would we be talking about implied vs realized vol on!
You need to delta hedge to isolate PnL to just the IV/HV spread and the literature is just as consistent in saying perfect delta hedging is impossible. Edit: Missed the last sentence. If you could trade say, variance swaps which are pure IV/HV plays, then reliably and accurately predicting future vol would guarantee profits.
I think you're talking about risk free arbitrage and I'm talking about abnormal returns. Simply consistently selling an overpriced asset, which options would be if realized vol was consistently lower than implied, would generate abnormal returns. Still curious as to your source though on the idea that realized vol is consistently lower than implied?
I'll give you that. I was just being academic. Must've been someone else I never said that. I only said if you could reliably predict when HV will be less than IV.
My bad, it was heiasafari with the "Realized vol is almost always lower than implied vol (except on rare occasions)" comment.