Using nothing but the data available in any given index, how do you derive a sigma from returns that's within 5% of implied volatility? Is it possible? I can get close to 10% using pivot means converted to rates and sectored ratios of a fiscal year, but other than that, I'ld imagine I'ld need a comparable array of data to couple and then interpolate. Anyone have experience in this? At this late stage, quants on the street _still_ compounding their rates scares me.