Selling delta 3-5 ES Puts with 40-57 days left

[Maverick74]

@Mav74. Do you have a minute to comment on my question in post #33 an #37 ? thanks

This is the only question I saw:

"The one thing you have not answered yet (and again I respect if you dont' want to but just say so that way I dont keep bugging you : ) is whether you always let your straddles go to expiry or whether you close it out at some point when it's gone too far ITM on one side? "

I never let options go to expiration long or short. And I trade all sorts of combinations of options not just straddles. Not sure what this information does for you but I don't subscribe to any kind of option orthodoxy.

 

[sss12]

This is the only question I saw:

"The one thing you have not answered yet (and again I respect if you dont' want to but just say so that way I dont keep bugging you : ) is whether you always let your straddles go to expiry or whether you close it out at some point when it's gone too far ITM on one side? "

I never let options go to expiration long or short. And I trade all sorts of combinations of options not just straddles. Not sure what this information does for you but I don't subscribe to any kind of option orthodoxy.

 

[sss12]

@i960 and Mav74...good info, thanks

 

[sle]

Mav, you can't deny that in modern day regulatory environment, there is clear alpha in selling some stuff (e.g. crash cliquets are crazy rich). The knowledge and understanding of what to sell is not easy to come by, though

 

[tradelosses]

as someone who has tried this, don't. this method sucks. Them little puts get big real fast. A put I sold 2 weeks ago at was at $1 and went to $3.6 this morning before closing at 2. Even with 2.5 weeks of theta burn on the 51 day options to my advantage, the losses still ballooned over the past 6 days. At day 51 I sold the 1780 put at 2.5. 2 weeks later it went at low as 1.5. At day 34, Tuesday, it was at 5. The time decay cannot keep up with the explosion in volatility.

I made some adjustments to the trade to take some of the gamma risk off. I sold the 1100-1300 170-day puts on the ES to capture more premium but reduce gamma. I also hedged by shorting SPXL (never ever short SPY or go long SDS. always hedge by shorting the 3x long etfs or buying their puts). 51 days is not enough time. The gamma risk is too high.

The high sharpe does NOT take into account path dependence. When your delta goes from 10 ($20,000 on the ES) to 27, you are hemorrhaging thousands of dollars in the process. it gets ugly so quickly.

 

[i960]

I made some adjustments to the trade to take some of the gamma risk off. I sold the 1100-1300 170-day puts on the ES to capture more premium but reduce gamma.

How is this reducing gamma on a trade where you're already short gamma from selling puts?

 

[ironchef]

as someone who has tried this, don't. this method sucks. Them little puts get big real fast. A put I sold 2 weeks ago at was at $1 and went to $3.6 this morning before closing at 2. Even with 2.5 weeks of theta burn on the 51 day options to my advantage, the losses still ballooned over the past 6 days. At day 51 I sold the 1780 put at 2.5. 2 weeks later it went at low as 1.5. At day 34, Tuesday, it was at 5. The time decay cannot keep up with the explosion in volatility.

I made some adjustments to the trade to take some of the gamma risk off. I sold the 1100-1300 170-day puts on the ES to capture more premium but reduce gamma. I also hedged by shorting SPXL (never ever short SPY or go long SDS. always hedge by shorting the 3x long etfs or buying their puts). 51 days is not enough time. The gamma risk is too high.

The high sharpe does NOT take into account path dependence. When your delta goes from 10 ($20,000 on the ES) to 27, you are hemorrhaging thousands of dollars in the process. it gets ugly so quickly.

Being there done that. Net negative profit after a year selling calls and puts. Small cumulative profits could not cover a few surprises. Rolling sometimes helped but mostly didn't.

After reading Maverick74's posts on short DOTM calls and puts, I now mostly avoid them.

How is this reducing gamma on a trade where you're already short gamma from selling puts?

Was he selling a 1100-1300 credit puts spread? But if so it is also short gamma? I am confused.

 

[sle]

In my limited opinion, the only persistent alpha in S&P vol is the overpriced crash risk. These days it is driven by rather stringent capital adequacy limits imposed by Basel III, combined with HF reporting requirements. I.e. if you are selling short-dated options that would only pay in an theoretical crash scenario, you are probably harvesting somebody's lunch. That somebody is most probably a vanilla MM desk at a Basel-compliant bank. The very same people that pay 60-70bp/a for S&P 500 85/75 daily crash cliquets or 1 pt vol spread for 2.5x capped/uncapped variance. They don't have a choice. Even the funds can't sell this stuff now since the big guys have some sort of crash limits from their PBs (again, due to Basel III)

So here is a theoretical strategy. Every Friday you wake up and load the full option chain for the next week SPX/SPYs (depending on your preference). Now, through the day, check where the lowest-strike nickel and dime bids are - if they are above some reasonable threshold, you should hit that bid. Usually you will find that you are selling some silly strikes - I think last Friday lowest nickel bid was 1250 at some point during the day (that's for 1 week expiration).

Obviously, for a strategy like this you primary risk is vega convexity, not price convexity. You are not afraid that these things will be ITM at expiration, but that their premiums will spike. Since there is no real hedge for vega convexity, you wanna manage your margin accordingly and not get too stingy. This said, I'd imagine a conservative approach in this can still yield very reasonable results for a private trader.

PS. There is also some alpha in the vega end of the curve, but taking advantage of it is very tricky.

 

[isotope]

In my limited opinion, the only persistent alpha in S&P vol is the overpriced crash risk. These days it is driven by rather stringent capital adequacy limits imposed by Basel III, combined with HF reporting requirements. I.e. if you are selling short-dated options that would only pay in an theoretical crash scenario, you are probably harvesting somebody's lunch. That somebody is most probably a vanilla MM desk at a Basel-compliant bank. The very same people that pay 60-70bp/a for S&P 500 85/75 daily crash cliquets or 1 pt vol spread for 2.5x capped/uncapped variance. They don't have a choice. Even the funds can't sell this stuff now since the big guys have some sort of crash limits from their PBs (again, due to Basel III)

So here is a theoretical strategy. Every Friday you wake up and load the full option chain for the next week SPX/SPYs (depending on your preference). Now, through the day, check where the lowest-strike nickel and dime bids are - if they are above some reasonable threshold, you should hit that bid. Usually you will find that you are selling some silly strikes - I think last Friday lowest nickel bid was 1250 at some point during the day (that's for 1 week expiration).

Obviously, for a strategy like this you primary risk is vega convexity, not price convexity. You are not afraid that these things will be ITM at expiration, but that their premiums will spike. Since there is no real hedge for vega convexity, you wanna manage your margin accordingly and not get too stingy. This said, I'd imagine a conservative approach in this can still yield very reasonable results for a private trader.

PS. There is also some alpha in the vega end of the curve, but taking advantage of it is very tricky.

Does https://en.wikipedia.org/wiki/Basel_III stand corrected? It claims that Basel III won't be implemented until 2019. I raise this question only out of skepticism of Wikipedia's quality of information.

 

[tradelosses]

How is this reducing gamma on a trade where you're already short gamma from selling puts?

the deltas are the same but the gamma is less because the duration is 170 days, not 30. As the market falls, these puts don't budge as much as the 30 day ones