I'm looking for data on this. The CBOE has an index, the Put Protection Index http://www.cboe.com/products/strategy-benchmark-indexes/put-protection-index but the problem is that the data of that index is bundled with the return of the S&P500, and even if you remove the S&P500 return from there (which I have done), you still can't know what was the % return of the premium invested because you don't know how expensive were the options at the time of purchase. I'm looking to build a monthly return series with this. All you can know from their data is that buying OTM puts loses money in the long-run but to me, this doesn't say much, if it produces big enough returns every once and a while, profits can be reinvested into crashing markets which produces a 'rebalancing return' which can offset the negative expectancy, not to mention in the improvement in risk adjusted metrics. I want to run a backtest to check all of this Anyone know where I can get some data on this? the CBOE has some SPX option data going back to 1990, but it costs $65,000 which is ridiculous
http://www.ivolatility.com/data/us-historical-option-data2.html They seem to have taken the prices off the page but it is very affordable. Edit from an old thread: Small amounts of data e.g just closing option prices for all strikes for one year is $3.60 per stock, $10 per ETF, $20 per index.
Re-check your cost numbers! Their (CBOE LiveVol) data (SPX option quotes) goes back to 2004. Their 1-min data cost less than $1K -- less frequent intervals is cheaper.
http://www.optionstack.com/ You can use this service to backtest option strategies using real price data, not synthetic or 'theoretical' option prices. Their historical option data goes back a couple of years so you can get a good idea what this type of "tail hedge" strategy would cost you in different volatility environments.
I have to disagree. -- It's better to look at constant tenor data as opposed to using constant maturity contracts for back testing. You can always add conservative execution assumptions while your stats are going to be more reliable. -- Unless they supply synchronized tick-level option data for each strike and the underlying (I really doubt that, since it looks like a shoe string startup) whatever execution assumptions are going to be borderline useless. -- Couple of decades is not enough for the types of studies he's trying to do, less so couple years (especially recently).
You're exaggerating. He's trying to get something comparable to the CBOE Putwrite index, he's not asking for a high-frequency intra-day tick-level backtest. Using SPX or SPY historical End-Of-Day option data will give him a very good idea of the cost of implementing a tail-hedge strategy for a S&P 500 equity portfolio. I have no idea why you insist on "synchronized tick-level option data for each strike" which is an absolutely ludicrous requirement for the question Daal asked. You can criticize optionstack.com all day long -- but what about giving an alternative and better suggestion at a similar price level?
Rolling x days or months as opposed to using fixed expiration as represented by the exchange traded options. This way, you have much broader sample space with respect to distribution realized by the underlying