Had worked well at high frequency, esp. wrt ETFs vs slightly dirty hedges. Have moved to a longer timeframe now .. with the privilege of slow execution, transaction fees, and own fkn capital at risk.. Sample size is too small to comment with any meaning presently.
Lots of variables. He'd sell the call (over the symmetrical put) on a flat surface due to financing, so it's moot. He sees a bigger premium. I'd rather sell a 4x1 ratio or PF (short one par put, short two par calls). I haven't traded CL actively since 2004, but if I did I would rather take a trade on CL-direction in gamma then in the vol-line. I doubt the persistence of the skew (or shape thereof) is going to change while I am short the var-swap.
I have not traded any commodities for a while by now, but lets take it to a general level. Say if you are trading implied vs realized, wouldn't variance make more sense since the skew is, in general overpriced vol-wise (and you are not running the risk of re-balancing your delta at the worst possible time)?
Sure, if you're arguing a discrete/dynamic hedging scenario. I was referring to an outright bet on vol or price. Which do you think closer approximates betapig's method?
Here is an update of that portfolio I posted earlier this week. FYI, Inception date was June 22, 2011. Profit is now at 3.113%, up from 1.725%. It seems like everything is working so far. http://www.betapeg.com/updated.jpg Any more comments, suggestions, or observations are welcome as long as they're respectful. Here is the previous screenshot in case you missed it. http://www.betapeg.com/P0sitions.jpg
Here is a suggestion to you - what do you think is the "historical" value of your portfolio? Actually, do it two ways. First, simply take the history of each asset and take the percentage returns of each asset over your horizon (e.g. if you have 25 business days take all 25 day returns) and calculate the payoff of your option for each day. Second, do the same while rescaling the returns by the ratio of preceeding N-day standard deviation to recent N-day standard deviation. The question you want to answer is - "am i simply underwriting tail events or extracting some sort of excess value?" Pay attention, however to the statistical significance given the small amount of data. E.g. If you only have 10 years of history for corn, looking at payoff of a 2-delta options is a bit hard. SPX, on the other hand, has about 100 years by now and (if you assume that the markets did not change much) you could have a reasonable historical estimate.
Well, best way of describing it would be "distributional arbitrage" so it combines both direction and volatility. To be honest, my issue with these guys is not that these strategies are totally wrong, I do believe that there is alpha in selling wings. It's that they (a) frequently do it as a mechanical system instead of actually evaluating the alpha in the trade (b) frequently think that they have invented something that nobody had thought of before and thus they are golden geniuses (c) do not pay much attention to risks and discount rare events (d) and, most importantly, rather quickly try to become "asset managers" or "teaching gurus" The combination of (b) and (d) is definitely the worst.