How to trade delta-neutral straddles if iv can be forecasted reasonably well?

[Phoenix_colossal]

This may be a dumb question. Say I have a model that can forecast the delta neutral straddle iv of one asset reasonably well, out of sample, with a horizon of 10~20 trading days. But the price of the underlying, or the price of the straddle can't be predicted. How should I trade it? I assume one would need to constant delta-hedge the position remain delta neutral.

For delta-neutral straddle (or just ATM option for approximation), assume simple BS formula, short duration, it can be reduced to (https://quant.stackexchange.com/que...l-approximations-to-the-black-scholes-formula)
call = put = StockPrice * 0.4 * volatility * Sqrt( Time )
So essentially I can forecast the ratio (iv) between the option price and underlying price. I have read some papers like this: https://www.researchgate.net/public...y_prices_an_application_to_options_on_the_DAX. They just constantly close/open new straddles depending on if tomorrow's iv forecast is higher/lower, which doesn't seem very practical for many assets due to slippage. Also this way it is not purely trading iv, still gets impacted by underlying's movement. Any suggestions? thank you.

 

[OptionsOptionsOptions]

• Long or short straddles?
• I assume delta-neutral is a clue - but option greeks are all gobbledygook to me.

 

[Bastion_01]

This may be a dumb question. Say I have a model that can forecast the delta neutral straddle iv of one asset reasonably well, out of sample, with a horizon of 10~20 trading days. But the price of the underlying, or the price of the straddle can't be predicted. How should I trade it? I assume one would need to constant delta-hedge the position remain delta neutral.

For delta-neutral straddle (or just ATM option for approximation), assume simple BS formula, short duration, it can be reduced to (https://quant.stackexchange.com/que...l-approximations-to-the-black-scholes-formula)
call = put = StockPrice * 0.4 * volatility * Sqrt( Time )
So essentially I can forecast the ratio (iv) between the option price and underlying price. I have read some papers like this: https://www.researchgate.net/public...y_prices_an_application_to_options_on_the_DAX. They just constantly close/open new straddles depending on if tomorrow's iv forecast is higher/lower, which doesn't seem very practical for many assets due to slippage. Also this way it is not purely trading iv, still gets impacted by underlying's movement. Any suggestions? thank you.

if you re long vol, PnL will roughly average vega *( future realized vol - Implied you paid for), vice versa for shorts. delta hedging frequency will determine your expected PnL vol

 

[kcgoogler]

This may be a dumb question. Say I have a model that can forecast the delta neutral straddle iv of one asset reasonably well, out of sample, with a horizon of 10~20 trading days. But the price of the underlying, or the price of the straddle can't be predicted. How should I trade it? I assume one would need to constant delta-hedge the position remain delta neutral.

For delta-neutral straddle (or just ATM option for approximation), assume simple BS formula, short duration, it can be reduced to (https://quant.stackexchange.com/que...l-approximations-to-the-black-scholes-formula)
call = put = StockPrice * 0.4 * volatility * Sqrt( Time )
So essentially I can forecast the ratio (iv) between the option price and underlying price. I have read some papers like this: https://www.researchgate.net/public...y_prices_an_application_to_options_on_the_DAX. They just constantly close/open new straddles depending on if tomorrow's iv forecast is higher/lower, which doesn't seem very practical for many assets due to slippage. Also this way it is not purely trading iv, still gets impacted by underlying's movement. Any suggestions? thank you.

Hello Phoenix,

I believe what you are looking for are vol swaps. But afaik they are not all that accessible to retail traders. Probably OTC if you have access to them somewhere. I could be wrong though..

Hth
-gariki

 

[Phoenix_colossal]

if you re long vol, PnL will roughly average vega *( future realized vol - Implied you paid for), vice versa for shorts. delta hedging frequency will determine your expected PnL vol

Thanks for the reply. From the PnL formula, does it mean that even if one can forecast IV perfectly, the PnL is still not deterministic because realized vol/vega is not known? Maybe I am missing something here. Is there a trade setup that can profit purely based on IV?

Here is one plot I am working on. The 2M delta-neutral straddle price on /CL, vs future close and IV. The straddle price is following the 0.4 x underlying x IV x sqrt(DTE) reasonably well. Say for period between July 08 to Jan 09, IV of the straddle is increasing (doubled), but the underlying is dropping fast, as a result the straddle price dropped a lot. So if I forecast perfectly that IV would rise in the 2nd half of 08, and long the delta-neutral straddle, I would be losing money on the straddle, and also incur losses on the delta hedge (since I would be increasing long futures in a declining market to delta hedge)? I feel like something is wrong in this analysis but don't know where...

 

[Phoenix_colossal]

Hello Phoenix,

I believe what you are looking for are vol swaps. But afaik they are not all that accessible to retail traders. Probably OTC if you have access to them somewhere. I could be wrong though..

Hth
-gariki

Thanks. That's what I am afraid of. I read about vol/variance swaps, seems like they are more of a "pure play" on vol, whereas the vanilla options for retails have path dependence on the underlying. I remember either CBOE or CME introduced vol/variance swaps for retails years ago, but remains dead with zero volume over the years?

 

[Bastion_01]

Thanks for the reply. From the PnL formula, does it mean that even if one can forecast IV perfectly, the PnL is still not deterministic because realized vol/vega is not known? Maybe I am missing something here. Is there a trade setup that can profit purely based on IV?

Here is one plot I am working on. The 2M delta-neutral straddle price on /CL, vs future close and IV. The straddle price is following the 0.4 x underlying x IV x sqrt(DTE) reasonably well. Say for period between July 08 to Jan 09, IV of the straddle is increasing (doubled), but the underlying is dropping fast, as a result the straddle price dropped a lot. So if I forecast perfectly that IV would rise in the 2nd half of 08, and long the delta-neutral straddle, I would be losing money on the straddle, and also incur losses on the delta hedge (since I would be increasing long futures in a declining market to delta hedge)? I feel like something is wrong in this analysis but don't know where...

View attachment 267869

vega is known, just do a web search for greek calculation formulas. It feels like you dont really understand how options are valued yet, it's best to master the fundamentals first.