I have been trading options spreads for my personal account, bear put / bull call depending on my views on specific stocks, for the past 1.5 years.

Last year my portfolio had a 58% return compared to a flat SP500 index. YTD 2012 I am almost on par.

While “beating” the index has been my objective so far, I am aware its not really a fair comparison on risk adjusted terms. Cant use standard deviation because of the non linearity in options payoffs.

Really struggling to identify an appropriate benchmark.

As you mentioned, % return on options does not make sense due to leverage. You should look at it in terms of option notional.

For instance, if you buy 1000 options (assume multiplier = 1, for simplicity) and the stock price is $100, you should compare this to a stock portfolio worth $100,000. The reason is that one your options are in-the-money, the payoff ratio is 1:1 with $100,000 in original portfolio value.

if i understand you right, a deep ITM option spread would have delta = 1 hence payoff 1:1 with $100k stock portfolio per your example.

however there are a couple of points that, in my opinion, differentiate from a pure stock portfolio:

the options spread has a finite life

even an ITM spread has a non-zero probability of returning 0 i.e. it can go OTM

realistically, it wouldnt be possible to generate excess returns unless one initiates a position with either an OTM spread or an ITM spread on a volatile stock. hence the risk of total capital loss is not insignificant (unlike a stock portfolio) until very close to the option end date.

Hmm. Not sure if either of those matter. Options might have finite life, but you can always roll the positions or treat the option payoffs like discrete PnL events. If the option expires OTM, you just lose the premium amount, which again, is just like a discrete payoff. For instance, if you have a $100k portfolio and you lose $5k on OTM options, that’s just like losing 5%.

The subjective part, in my opinion, is what capital base to assume. For a risk adjusted basis, I would suggest the method above. However, you can add leverage to any portfolio, so it’s not clear that this method is always appropriate.

It’s true that Standard deviation doesn’t work so well. You might try downside deviation and maximum drawdown measures for risk, though there are the obvious representativeness problems.

Another way to look at risk is to figure out what scenario is most destructive to your portfolio and then try to figure out how often that scenario has occured or seemed likely to occur.