Just a random idea that probably everybody has had at some point, so it's probably not rocket science Selling puts on individual stocks from a certain index (atm or slightly otm). Then taking lets say, 10-30% of those returns to buy long puts on that same index. Looks good in theory, am i missing anything here in a real life scenario? I have nothing to backtest this with.
This is a reasonable idea in my eyes but you'll want to consider and be conscious of what you're actually hedging. Perhaps you're familiar with the CAPM model but if not it could be a good thing to get acquianted withA as it will help you quantify the market and beta exposures of each leg and therefore what specifically you are and are not hedging.
I tried naked puts at some point, but the PNL was pretty random, if not negative. Didn't try dispersion trade though - and at the moment I'm too busy with several ideas to have time for this. Maybe a platform to backtest options would be a success. Problem is I'm doing Java programming, backtests are intimely related to the architecture of the application I wrote and even so, every new idea seems to need either a refactor or some very creative hack. Programmig is hard...
You are buying correlation premium (the opposite of dispersion) and your strategy most probably has negative expected value unless you apply some other filters.
That might very well be, as nothing has been thought out very well and/or backtested. I would not try to hedge 100% though, just some long puts to dampen the volatility a bit. So it would pretty much look like naked puts with a bit less volatility, and lower returns.
Your index puts would help with a (small) degree of market risk, but not the specific risk of the company. So I don't get the point. That's expensive insurance that doesn't cover your biggest risk. While the market does have some impact on the movement of just about any stock, it pales in comparison to the company risk in the stock.
I don't mind it. I think if you are short puts in enough stocks it helps diversify the company-specific risk. The vol on an index vs. its components also changes, so this is the kind of trade that could work out much better in some index / markets than in others. Ideally, you would run some numbers on vol of an indices vs. it's components for different ETFs and see where you are getting the best value. As with any short put strategy, if you are short a lot of puts relative to your account size in one company you could run into trouble.