Yes there is! The code is already written and the backtest is complete, but I just want to cover one more idea this week, then this one will be next on the docket.
By the way I noticed that NDX options have a very wide bid/ask spread, compared to the SPX ones.
It would be nice to have a place to share thoughts and experiences about your ideas explained in these posts. Have your ever thought about creating a discord channel (or a telegram one) so your followers can stay in touch and help each other?
If you submit your orders a few (>=~$10) dollars away from the midpoint, you’ll get filled quickly. Executions will almost always be less than the midpoint (just the nature of retail brokerages), but the liquidity is there.
I appreciate your thoughts on considering getting some kind of community going! I’ll definitely be looking into this, I will try to do some polling in my next post to see what the best steps forward are. It’ll likely end up being a Discord, but I want to understand demand and preferences, things like that, first.
I finally managed to code this system with the interactive brokers API. Today I'm going to test it for the first time, with these last additions. So I will be working on NDX and will be using the 80% rule to close the risky legs.
Initially I will supervise the algorithm, that just triggers Telegram notifications when an event occurs. After some initial testing, I would like to let it place the orders itself, to close the legs when needed.
This sounds really exciting, I’m thrilled you were able to integrate this with your API and developed a notification system for it!
It’s a great year for this strategy and I’m sure you’ll come out ahead, just make sure to really stay firm on the 80% cutoff (or the cutoff you personally find more effective).
Please keep me updated with your findings, good luck!
Regarding entry timing, every day I would wait the first 90 minutes of the trading session before entering, to avoid the initial high volatility. Do You think it is a good idea or do You have other suggestions regarding entry timing?
If you have your system automated and it doesn’t need you to continuously watch the probability, then that doesn’t sound bad!
However, the time value of the option deteriorates fastest in the final hours of the session; when you sell the option early in the day, it loses value very slowly before picking up speed. So you would get a higher premium, but you would only actually see profits towards the final end of the session.
Additionally, leaving more hours in the day changes the volatility range of the system; so you would have to sell a volatility range of, say, 1.5% (up from usual 0.7%). From a risk standpoint this is fine as the range is generally accurate, but on an adjusted basis the premiums are about the same because this is the option-implied, 1 standard deviation volatility. So the payoff for 1 standard deviation of risk will usually remain the same (5%).
So, if you generally get $50 for selling a 0.7% range in the last 2 hours, then you’d get just a slightly larger amount for selling a 1.4% within the first 2 hours.
I personally suggest setting things up in a way that spends as little time in the trade as possible. From my experience, spending hours eyeing the probability gets tedious and unproductive after the first week, so the less time spent doing that, the better. This is paired with the benefit of there being less time for adverse events to materialize.
Trading early in the day increases the risk that if you hedge one side, but since there’s so much time left, price reverts and then you have to hedge the other side, resulting in a double-loss.
So in sum, let the volatility of the day do whatever it does, then get paid for collecting the last bit of it. The probability of high vol coming in the first 5 minutes is just as high as the probability of high vol coming in the last 5 minutes. We don’t know what that probability is, but it is easier to manage in the last 2 hours when the contracts are worth little, than in the first 2 when the contracts are still often in the hundreds.
Hey is there a plan to re-visit this and hook it into the polygon API?
Hi there,
Yes there is! The code is already written and the backtest is complete, but I just want to cover one more idea this week, then this one will be next on the docket.
Awesome thank you!
Hi,
first day closed with a profit :)
By the way I noticed that NDX options have a very wide bid/ask spread, compared to the SPX ones.
It would be nice to have a place to share thoughts and experiences about your ideas explained in these posts. Have your ever thought about creating a discord channel (or a telegram one) so your followers can stay in touch and help each other?
Great to hear that you’re already seeing positive results with it!
Towards the end of the post, I mentioned how those traditional metrics are misleading signals of liquidity for index options: https://www.nasdaq.com/articles/index-options-are-more-liquid-than-the-screens-indicate
If you submit your orders a few (>=~$10) dollars away from the midpoint, you’ll get filled quickly. Executions will almost always be less than the midpoint (just the nature of retail brokerages), but the liquidity is there.
I appreciate your thoughts on considering getting some kind of community going! I’ll definitely be looking into this, I will try to do some polling in my next post to see what the best steps forward are. It’ll likely end up being a Discord, but I want to understand demand and preferences, things like that, first.
Thanks a lot, I didn't understand that liquidity fact but I now got it.
If You need additional help to setup and manage the community space, I'm available and feel free to ask!
Hi,
thanks a lot for this follow up.
I finally managed to code this system with the interactive brokers API. Today I'm going to test it for the first time, with these last additions. So I will be working on NDX and will be using the 80% rule to close the risky legs.
Initially I will supervise the algorithm, that just triggers Telegram notifications when an event occurs. After some initial testing, I would like to let it place the orders itself, to close the legs when needed.
Hi Marco,
This sounds really exciting, I’m thrilled you were able to integrate this with your API and developed a notification system for it!
It’s a great year for this strategy and I’m sure you’ll come out ahead, just make sure to really stay firm on the 80% cutoff (or the cutoff you personally find more effective).
Please keep me updated with your findings, good luck!
Thanks a lot.
Regarding entry timing, every day I would wait the first 90 minutes of the trading session before entering, to avoid the initial high volatility. Do You think it is a good idea or do You have other suggestions regarding entry timing?
If you have your system automated and it doesn’t need you to continuously watch the probability, then that doesn’t sound bad!
However, the time value of the option deteriorates fastest in the final hours of the session; when you sell the option early in the day, it loses value very slowly before picking up speed. So you would get a higher premium, but you would only actually see profits towards the final end of the session.
Additionally, leaving more hours in the day changes the volatility range of the system; so you would have to sell a volatility range of, say, 1.5% (up from usual 0.7%). From a risk standpoint this is fine as the range is generally accurate, but on an adjusted basis the premiums are about the same because this is the option-implied, 1 standard deviation volatility. So the payoff for 1 standard deviation of risk will usually remain the same (5%).
So, if you generally get $50 for selling a 0.7% range in the last 2 hours, then you’d get just a slightly larger amount for selling a 1.4% within the first 2 hours.
I personally suggest setting things up in a way that spends as little time in the trade as possible. From my experience, spending hours eyeing the probability gets tedious and unproductive after the first week, so the less time spent doing that, the better. This is paired with the benefit of there being less time for adverse events to materialize.
Trading early in the day increases the risk that if you hedge one side, but since there’s so much time left, price reverts and then you have to hedge the other side, resulting in a double-loss.
So in sum, let the volatility of the day do whatever it does, then get paid for collecting the last bit of it. The probability of high vol coming in the first 5 minutes is just as high as the probability of high vol coming in the last 5 minutes. We don’t know what that probability is, but it is easier to manage in the last 2 hours when the contracts are worth little, than in the first 2 when the contracts are still often in the hundreds.
thanks for the complete answer!
so basically just using the last 2-3 hours is enough to collect a decent credit, while avoiding to be exposed to the market for a long time