Is it possible to profit from divergence in deltas ? What should I take into account ? Have you got insight ? The strategy would be something like : - Generate possible outcomes by bootstrapping historical data - Buy undervalued delta (Delta < P(Bootstrap)) - Sell undervalued delta (Delta > P(Bootstrap)) For exemple, I take one year of XYZ historical data. I randomize it and create different cumulative returns. I get a PDF of possible outcomes for different expiration dates. From there I see some XYZ deltas for a given expiration diverges. I buy delta which are undervalued according to my probability distribution. And sell those which are overvalued...
It's a weird way to phrase it, but yes, you can do this type of stuff... Obviously, as always, past performance doesn't necessarily predict future results.
I'm going to say no. I'm not sure how you can arb a Greek. The Greeks are informational. The inputs into the model like IVol, Interest rates/Dividend flows are the inputs that can be right or wrong.
Find a Greek who doesn't know anything about derivatives and trade a lot with him Though given how wiley that country has been, he might in the end, arb you.
i think your thinking more about dispersion... trading a basket of undervalued goods against a basket of over valued goods... like extracting out the undervalued stocks from a ETF and selling the overvalued and buying the undervalued.. so your not taking outright stock risk... its a creation of different optionality then the outright.. with its own set of risks..
That's it. However I am trying it manually for the moment. It's time consuming so I'll try to automate and backtest it. Then I could be able to implement it on a large scale. I don't really like the concept of arbitrage. Especially for a retailer like I am. Going to see what I can do. Just wanted some feedbacks.
A bit of advice you can take for what it's worth based on this and several of your other recent posts. In general, options and other derivatives are efficiently priced based on the likelihood of the underlying events occurring up on which they are based. You aren't going to be able to "arbitrage" anything, if you're using generally accepted definition of that word in finance which is entering a risk-free transaction that involves buying an underpriced and selling an overpriced security that represent the same underlying and will certainly converge to the same value. If you believe that the market has the volatility wrong, then you can take a bet on that using derivatives. But don't tell yourself you're arbing anything, you're taking a directional bet, pure and simple. Nothing wrong with directional bets, unless you're making them while fooling yourself into believing that you're arbing.
welll even in the most strict definition of the arbing sense there is risk.. wheat in chicago is a different product then wheat ANYWHERE else... its a different product! if it was completely and absolutely fungible it would occupy the exact same locations risks etc.. so to me there is definitly different degrees of riskful arb... no such thing as risk"less" and yes trading delta hedged anything isn't arb its a bet on the direction of vol.. trading baskets against eachother is no different... or even silver on one exchange to silver on another exchange, they are different and carry there own riskss....
Agree. I don't care about EMH What I care about is my P&L. I don't want to bend the market. Just find a way to make money from it. So if that stuff works, fine. What's the payoff ? I don't like arbitrage since it looks like selling options. Not that kind of guy. I prefer to buy option. As far as efficient the market can be. There will always be some kind of opportunity. Black Swans ... Order Flow ... Positive expectancy... However Risk free is for economists. Traders know there ain't free lunch. I take the market as it is. Build model, strategy, that work or fail.
Options markets are too efficiently priced for there to be any arbs such as this to actually work. You're wasting your time, but don't let me stop you from giving it a go.