volatilityhq.com Official Thread

[Djtux 548]

Hi all,

I see there are multiples questions about the website, i figure i will create a dedicated thread to answer all of them.

 

Link to the website : https://www.volatilityhq.com/payments/confirm/1/?coupon_code=SO502411 (you need to create an account first).

For members of this forum, there is a 50% forever coupon is SO502411 and must be redeemed by Nov 30 2024. The 50% discount will be applied for as long as you keep your subscription (please make sure to update your credit card if it expires).

 

FAQ:

• How can i contact you for support questions ?

Quote

Sure, send me an email to support(at)volatilityhq.com. Just replace the (at) by @.

• How much do you charge ?

Quote

The regular monthly price is $99.99/month but can be increased at any moment so be sure to subscribe today. There is a 50% coupon (see above).

• Why are you not using paypal ?

Quote

We use stripe.com to process the payment. Our website never sees your credit card number : they are sent directly from your browser to stripe.com. Stripe is a very reputable company that is used by multiple companies like Kickstarter, Lyft (see there list https://stripe.com/us/customers)

• How can i cancel the subscription ?

Quote

You can cancel your subscription on www.volatilityhq.com billing profile. You just need to click the cancel button, confirm and you are all set. The credit card won't be charged anymore, and you can still enjoy the website until the end of the current month.

• Is there a documentation ? I'm not clear how to interpret the charts.

Quote

Yes, the charts are not easy to interpret. We plan to write a more detailed documentation but in the mean time, if you are a SO subscriber, you can refer to https://steadyoptions.com/forums/forum/topic/1393-tools-to-analyze-the-earnings-trades/.

Basically the RV (relative value) chart displays the RV of a straddle or a calendar, and helps determine how cheap/expensive a strategy is compared to previous cycles.

RV straddle = straddle price / stock price

RV calendar = calendar price / stock price

• Why is the website not free anymore ?

Quote

We believe that the website provides value in its current form, and keeping the website free would make it unsustainable. We hope that you understand and will support us. We have new features planned as well in the future.

• I would like to see X feature to be implemented ? Can you do it ?

Quote

If this is something that can benefit everyone and is technically doable, then yes, please share your idea and we will try to plan it.

• Where are you getting your data ?

Quote

Because this is a public thread, i prefer not to share or discuss that information here.

• Can i share a screenshot of the charts from https://www.volatilityhq.com on SteadyOptions forum ?

Quote

Yes of course, but please no automated scraper :).

• Do you charge sales tax ?

Quote

If you live in New York state (in the US), i have to charge sales tax.

 

Edited November 8 by Djtux

[Djtux 548]

[reserved for later]

 

Testimonials :

• https://steadyoptions.com/forums/forum/topic/3969-september-18-22-2017-trading-candidates/?do=findComment&comment=81233
• https://steadyoptions.com/forums/forum/topic/3884-volatilityhq/?do=findComment&comment=80003
• https://steadyoptions.com/forums/forum/topic/2346-how-we-trade-straddles-and-strangles/?do=findComment&comment=86786

Edited November 11, 2017 by Djtux

[Optrader 128]

Is it possible to change the straddle expiration week ? Rather than the default which is weekly expiration ?

[Optrader 128]

Same goes for calendar.

can we change both long and short expiration week .

default can be what it is now.

 

 

[Djtux 548]

4 minutes ago, mukundaa said:

Is it possible to change the straddle expiration week ? Rather than the default which is weekly expiration ?

There is the option "Minimum number of business day between earning date and straddle expiry date" which is 0 by default.

If you select 5, then the tool will select the expiry date such as there are at least 5 business days between the earnings date and the expiry date.

So instead of the weekly expiry just after the earnings date, you will get the expiry 1 week after.

On the right you see the earning date selected for each cycle to double check.

2 minutes ago, mukundaa said:

Same goes for calendar.

can we change both long and short expiration week .

default can be what it is now.

For calendar, there is the option "Number of weeks between the calendar legs".

By default, it will select the expiry date just after the earnings date for the short leg, and the next monthly expiry date for the long leg.

The option can be changed between [1,2,3,4], and will specify the number of weeks between the short leg and the long leg. The short leg will always be the expiry just after the earnings date, you can't change that.

And you can double-check the expiry dates chosen for each cycle on the right of the chart.

[Optrader 128]

Great.

Thanks.

It is really wonderful site. But not sure how many customers would you get outside SO who would be interested in saddle prices/calendars.

[Djtux 548]

1 minute ago, mukundaa said:

Great.

Thanks.

It is really wonderful site. But not sure how many customers would you get outside SO who would be interested in saddle prices/calendars.

Thanks !

With the current version of the website, it's going to be 0 potential customers outside of SO.

But that's okay, it's better to be laser-focused initially to do 1 thing well, rather than be a jack of all trade.

But this is just the beginning, i plan to add more features.

[RapperT 336]

some underlings have little data.  How do you choose which stocks you provide data for and will this bucket of stocks increase in the near future?

Also, kind of an odd response about the data.  Why would it matter that this is a public forum?  It's not a big deal, im just curious.

 

 

[Djtux 548]

3 minutes ago, RapperT said:

some underlings have little data.  How do you choose which stocks you provide data for and will this bucket of stocks increase in the near future?

Also, kind of an odd response about the data.  Why would it matter that this is a public forum?  It's not a big deal, im just curious.

The website supports all US stocks ticker where the total daily option volume > 100 contracts.

Do you have some ticker in mind ?

For the public forum, mostly because it's cached on google and web.archive.org.

[RapperT 336]

i checked 3-4 last night in which there would be half of one price line or something similar.  I can get you some tickers.  I was just using optionslam calendar with some basic volume screens and then checking them on your site

Edited August 20, 2017 by RapperT

[RapperT 336]

maybe it has to do with days until earnings?  some have incomplete data if its more than a couple weeks out.  KMX one example

[Djtux 548]

6 minutes ago, RapperT said:

i checked 3-4 last night in which there would be half of one price line or something similar.  I can get you some tickers.  I was just using optionslam calendar with some basic volume screens and then checking them on your site

If you see ticker like that, send it to me so i can double-check. I can see 3 possibilities :

• a bug in my code
• the ticker data was filtered out because of the rule to ignore ticker with daily option volume <= 100 contracts
• the earning date is far in the future (like T-14 or more, not sure of the exact threshold). Let's say the weekly option that expires just after the earning date is selected for a given cycle. It is possible that at T-14 or T-20, the weekly option does not exist yet.

[Djtux 548]

5 minutes ago, RapperT said:

maybe it has to do with days until earnings?  some have incomplete data if its more than a couple weeks out.  KMX one example

Just checked on TOS thinkback, the 2017-06-23 weekly option was first trade on 2017-06-08.

That explains why the 2017-06-20 cycle does not appear before T-8.

You can see that for the cycles before 2017-04-05, there were no weekly.

I'm not sure what's the best way to handle stocks where for the older cycles there were no weeklies.

I though at the minimum displaying on the right of the chart the number of business days between the earning date and the expiry date to quickly see if there is an abnormal gap (because of no weekly options available). Another possibility would be to add an option to force to use only monthly options for the straddle.

Any suggestions is welcome !

 

You could try to play with the "Minimum number of business day between earning date and straddle expiry date", but it's not perfect.

I believe Yowster noticed that issue in some other thread.

[RapperT 336]

makes sense.  I'll keep you posted if i notice anything else.  great tool.

I really like being able to see out to t-20 (or something similar) when utilizing straddles hedged with short options

 

 

[Khonsu 9]

Hi @Djtux, would it be possible to share your list of planned enhancements?

Are there any plans to provide the value of an ATM straddle bought X days out, as that specific straddle changes value leading up to earnings? As opposed to the current state which displays the value of an ATM straddle, recentered on the underlying, each day. Hopefully that makes sense.

[krisbee 1,965]

35 minutes ago, Khonsu said:

Hi @Djtux, would it be possible to share your list of planned enhancements?

Are there any plans to provide the value of an ATM straddle bought X days out, as that specific straddle changes value leading up to earnings? As opposed to the current state which displays the value of an ATM straddle, recentered on the underlying, each day. Hopefully that makes sense.

IMO, it'd be complex to show it in the charts. There is a workaround for this, you can CHECK mark the price movements in straddle, it shows how the price moved from the first day till option expiry.

You'll get an idea about the gamma gain your straddle might have produced.

[Khonsu 9]

1 hour ago, krisbee said:

IMO, it'd be complex to show it in the charts. There is a workaround for this, you can CHECK mark the price movements in straddle, it shows how the price moved from the first day till option expiry.

You'll get an idea about the gamma gain your straddle might have produced.

Agreed, it'd certainly be a very complicated algorithm to write, just thought it'd be a neat way to essentially combine price movement with the regular straddle RV chart to get the most accurate  possible measure of straddle performances leading into earnings. I do use the price change charts in addition to the RV charts to get a rough sense of potential for gamma gains.

[Djtux 548]

4 hours ago, Khonsu said:

Hi @Djtux, would it be possible to share your list of planned enhancements?

In my opinion for a SAAS (software as a service) like https://www.volatilityhq.com, there are 3 things in order of priority (sorry if it's too technical) : 

1. Uptime, performance, and up-to-date data : now that i launched the website officially, it leaves me more time to focus on things like performance : transition from in-memory cache on each server to some cache like Redis or Memcache (AWS Elasticache). Uptime is important as you don't want the service down, especially now that i charge members, so i have a plan to do things in the backend to provide higher-availability. Up-to-date data : i also need to take care of the data each day, even if it's automated, i feel i can improve that.
2. Bug fixes : self-explanatory, if there are bugs, i need to fix that first
3. New features : i have many features in mind to improve the service even more, but i'm interested to know what would be useful from the members. That would help guide the effort. I also have to keep in mind that there are some quick wins as well that is not too hard to code but provide business value. As for an exact list of enhancements, i was thinking about implementing a vote+comment system on the website to get feedbacks on what's important. But for some possible enhancements (keep in mind i might be talking about multi months development projects) :

• Earnings calendar / scanner : that could include a way to more easily see if a strategy is cheap or not (https://steadyoptions.com/forums/forum/topic/3724-may-22-26-2017-trading-candidates/?do=findComment&comment=72791) like an RV rank metric. That way, you have a table with each row being a ticker, and in columns you would get an RV rank today, and an RV rank between the RV today and the average at T=0 for example. From the table, you could see which ticker to focus the analysis, and check the RV chart like now. That's a big project technically, because i need to setup some worker servers to be able to process many tickers in the background.
• Export csv data from the chart (https://steadyoptions.com/forums/forum/topic/3769-discussion-unofficial-trade-ideas/?do=findComment&comment=78849)
• Make the website mobile friendly
• Be able to change the number of cycles and be able to go further back in time (let's say backtest the last 2 or 3 years)
• Handle the case when earning date is uncertain
• Handle the straddle/strangle strategy : i'm not completely satisfied with the RV charts to access if a stock is a good candidate for a straddle+strangle strategy. There is some research to do on how to analyze a stock to see if it's a candidate or not. If you guys have ideas on how to do that...
• Use intraday data to compute the RV charts : maybe something like 30 min or 60 min intraday : that's about 6 to 13 points per day. That also means that 6 to 13 times the size to store the data in database. And the data is not cheap. But i really want to do that at some point.

The list is not exhaustive.

Edited August 23, 2017 by Djtux

[Djtux 548]

3 hours ago, Khonsu said:

Are there any plans to provide the value of an ATM straddle bought X days out, as that specific straddle changes value leading up to earnings? As opposed to the current state which displays the value of an ATM straddle, recentered on the underlying, each day. Hopefully that makes sense.

I think someone ask me for that already, i can't seem to find the post.

To me it makes perfect sense and it complements the RV chart.

In my mind the user interface would have a slider (it could go from T-20 to T-1 for example) to select which X to buy the straddle, and then the chart would be updated to show the RV of that specific straddle. You could see the RV or the return (return might be useful as well).

I saw in some post that the "Show stock price change" could be improved to support a relative stock price instead of an absolute price change. I think that would be useful too.

 

I don't have any timeline on any of this though.

[Khonsu 9]

49 minutes ago, Djtux said:

In my opinion for a SAAS (software as a service) like https://www.volatilityhq.com, there are 3 things in order of priority (sorry if it's too technical) : 

1. Uptime, performance, and up-to-date data : now that i launch the website officially, it leaves me more time to focus on things like performance (transition from in-memory cache on each server) to some cache like Redis or Memcache (AWS Elasticache). Uptime is important as you don't want the service down, especially now that i charge members, so i have a plan to do things in the backend to provide higher-availability. Up-to-date data : i also need to take care of the data each day, even if it's automated, i feel i can improve that.
2. Bug fixes : self-explanatory, if there are bugs, i need to fix that first
3. New features : i have many features in mind to improve the service even more, but i'm interested to know what would be useful from the members. That would help guide the effort. I also have to keep in mind that there are some quick wins as well that is not too hard to code but provide business value. As for an exact list of enhancements, i was thinking about implementing a vote+comment system on the website to get feedbacks on what's important. But for some possible enhancements (keep in mind i might be talking about multi months development projects) :

• Earnings calendar / scanner : that could include a way to more easily see if a strategy is cheap or not (https://steadyoptions.com/forums/forum/topic/3724-may-22-26-2017-trading-candidates/?do=findComment&comment=72791) like an RV rank metric. That way, you have a table with each row being a ticker, and in columns you would get an RV rank today, and an RV rank between the RV today and the average at T=0 for example. From the table, you could see which ticker to focus the analysis, and check the RV chart like now. That's a big project technically, because i need to setup some worker servers to be able to process many tickers in the background.
• Export csv data from the chart (https://steadyoptions.com/forums/forum/topic/3769-discussion-unofficial-trade-ideas/?do=findComment&comment=78849)
• Make the website mobile friendly
• Be able to change the number of cycles and be able to go further back in time (let's say backtest the last 2 or 3 years)
• Handle the case when earning date is uncertain
• Handle the straddle/strangle strategy : i'm not completely satisfied with the RV charts to access if a stock is a good candidate for a straddle+strangle strategy. There is some research to do on how to analyze a stock to see if it's a candidate or not. If you guys have ideas on how to do that...
• Use intraday data to compute the RV charts : maybe something like 30 min or 60 min intraday : that's about 6 to 13 points per day. That also means that 6 to 13 times the size to store the data in database. And the data is not cheap. But i really want to do that at some point.

The list is not exhaustive.

Appreciate the thorough response! I actually work with engineers and product at a SaaS company myself so makes total sense to me. 

A simple voting and comment system for determining priority on enhancements sounds like an excellent way to capture feedback given I assume a very lean dev team, especially considering I'm sure the subscribers will all make for a helpful community.

I like how the roadmap looks and it has probably convinced me to subscribe.

[RapperT 336]

@Djtux agree 100% on intraday data.  Hopefully that becomes a reality.

 

Have you checked Quandl price wise compared to your current source.

[Yowster 9,174]

18 hours ago, Khonsu said:

Hi @Djtux, would it be possible to share your list of planned enhancements?

Are there any plans to provide the value of an ATM straddle bought X days out, as that specific straddle changes value leading up to earnings? As opposed to the current state which displays the value of an ATM straddle, recentered on the underlying, each day. Hopefully that makes sense.

A couple of comments related to this:

• For me the RV chart is the most valuable, and always will be.   IV (and therefore RV) changes have closer to the same pattern each earnings cycle for the most part (more predicable).   Stock price changes leading into earnings is much less predictable and just because a stock moved a lot leading up to earnings in one cycle doesn't necessarily mean it will this time.   Entering trades at good RV levels with short strangle sales covering/exceeding the typical RV decline heading into earnings days is the top criteria for trade entry, IMO.
• The tool already has the option to display the stock price movement for prior cycles.
• If there is another chart of straddle performance tracking a straddle at price xxx over time, then the results can be misleading at times:
  • I know the intent is likely to track gains due to stock price changes (gamma), but when looking at a chart like this straddle changes due to IV changes (vega) would not be readily apparent.  If the trade was entered at higher IV, then straddle could show a loss even though stock price moved.  Conversely, if trade was entered a good RV then straddle could show vega gain even though stock price did not move much.  And vega could be an overall market thing and not stock-specific (so not repeatable from one cycle to the next).
  • Bottom line for me is that having a chart showing past straddle price changes over time is fine... but not being able to tell how much of the change is due to gamma, vega and theta makes it not very useful to me for making a trade entry decision.
  • Having such a chart certainly won't hurt (the more data the better), but the RV chart is the most important for trade entry decisions with the other charts providing more background.

 

[Optrader 128]

When do you update the charts

like todays data. Do you do end of day ? At what time ?

 

[Djtux 548]

16 minutes ago, mukundaa said:

When do you update the charts

like todays data. Do you do end of day ? At what time ?

 

It's between 5:00pm and 6:00pm Eastern time. The process is automated.

Right now (5:33pm eastern) the charts have the EOD data for today (Aug 28th 2017).

[RapperT 336]

On 8/23/2017 at 7:56 AM, Yowster said:

 

• If there is another chart of straddle performance tracking a straddle at price xxx over time, then the results can be misleading at times:
  • I know the intent is likely to track gains due to stock price changes (gamma), but when looking at a chart like this straddle changes due to IV changes (vega) would not be readily apparent.  If the trade was entered at higher IV, then straddle could show a loss even though stock price moved.  Conversely, if trade was entered a good RV then straddle could show vega gain even though stock price did not move much.  And vega could be an overall market thing and not stock-specific (so not repeatable from one cycle to the next).
  • Bottom line for me is that having a chart showing past straddle price changes over time is fine... but not being able to tell how much of the change is due to gamma, vega and theta makes it not very useful to me for making a trade entry decision.
  • Having such a chart certainly won't hurt (the more data the better), but the RV chart is the most important for trade entry decisions with the other charts providing more background.

 

I think its a good discussion and an interesting point.

rather than track historical straddle price over time, we could construct an artificial perfectly delta neutral straddle and forecast price into earnings based on historical vol.  Your gamma problem is solved and this absolute value is really what we are trying to get at.   That would be some cool data to look at.  DJ and I have chatted a bit about this behind the scenes

[jbthree 6]

I just noticed text that indicates whether the earnings release date has been confirmed or not. I love it!

[Djtux 548]

1 minute ago, jbthree said:

I just noticed text that indicates whether the earnings release date has been confirmed or not. I love it!

Glad you like it, and thank you for being a subscriber, very much appreciated.

Just the standard disclaimer : although i try to give an confirmed/unconfirmed status, it's always best to double-check with the official investor relations website of the stock you are looking at. That's the only 'official' source of information.

[Djtux 548]

Some upgrades this weekend :

• confirmed/unconfirmed status from several different providers
• i now have the entire option historical data : there is no filtering anymore as i used to do. The immediate effect is that by default there are 8 historical cycles (2 years) instead of 6 before.

Right now you could change the 'start date' to include more history but i haven't tested that much.

[Djtux 548]

There are some issues with the EOD data for 09/14/17, i'm waiting for an answer from the provider. Sorry about that.

Edited September 15, 2017 by Djtux

[wr200m 22]

I think provider just got it ready at 10:48pm

Edited September 15, 2017 by wr200m
removed src

[Djtux 548]

4 minutes ago, jbthree said:

Thanks, @greenspan76. I haven't analyzed it in depth either. But I would love it if @Djtux could include a chart that showed %change in RV on the y-axis with the same X axis (time). That way, we could also look for RVs that, on a percentage basis, decline by less than we receive from our hedges. This chart would also help us better judge the optimal entry time.

I'm answering this point on this thread as it's more related to volatilityhq.com.

I see 2 ways to interpret your idea and i'm not sure which one is the more useful :

1. Given 1 stock, for each earning cycle, instead of showing the RV, i will divide the RV for each T by some reference RV, for example the RV at T-10. That reference could at first be selected at the current T-X date.
2. Or do you want for each date, the % change in RV between T-X-1 and T-X (% change in RV between the previous and and the next) ?
3. Or you had something else in mind ?

Seems that 2. is a little bit less useful.

If anyone else has an opinion, please feel free to share, it will help.

 

[jbthree 6]

I think it would be useful to see, on a percentage basis, declines between the current date and expiration. If RV tends to decline between 10% and 15% between now and expiration, but I know I can sell 3 cycles of strangles at 6% each for 18%, the trade would be one to consider now, regardless of how expensive the straddle is relative to the average.

 

But if I see that the % gained on the strangles will not cover the % change in RV in previous cycles now, I may see that, for example %gained on straddles > %decrease in RV in previous cycles, if I enter the trade 1 week before earnings release.

I would love to hear you think, as well as @RapperT,  @greenspan76 ,  @Yowster and others think.

Edited September 18, 2017 by jbthree

[Djtux 548]

@jbthree Here a small quick proof of concept i just added to volatilityhq. It's not enabled by default, and you need to click on "Show RV chart in %" option for straddle.

 

I could potentially move the reference date to the latest date (T-5 in the chart below for example) for the current cycle. Note that there are some cycles where the expiry is not quoted as early as the start of the chart, so some cycles are ignored.

 

Edited September 19, 2017 by Djtux

[jbthree 6]

Wow! Thanks! Would it make more sense to see the percent decreases from the current day (in this case -5)? For example, 2017-06-20 decreases about 10% from -5 to -1.  So does the average. In this case a 6% hedge wouldn't cover everything. The stock would have to move.

What do you think?

[Kim 7,943]

Yes, was going to ask the same question. Default start date is the current date, and yet the chart shows few days before the current date.

[Djtux 548]

18 hours ago, jbthree said:

Wow! Thanks! Would it make more sense to see the percent decreases from the current day (in this case -5)? For example, 2017-06-20 decreases about 10% from -5 to -1.  So does the average. In this case a 6% hedge wouldn't cover everything. The stock would have to move.

What do you think?

Yep, i can look into changing that. Probably more intuitive to have the current date as the reference date for the RV in %. I have to check if there isn't any issues doing that.

2 minutes ago, Kim said:

Yes, was going to ask the same question. Default start date is the current date, and yet the chart shows few days before the current date.

Yes i have not documented the 'start date' box, but it's actually the start date of the range of earning dates considered. For the moment, i have the EOD data but i don't have the full history of earning dates to go that far.

The idea would be : by default, the start date is 2 years before today's date (to have roughly 8 earning cycles), but in the future it could be changed to 3 years for example.

 

As for why the chart always shows few days before the current date, i personally like to see a bit of context, and that number of days shown before the current date is not configurable. Do you guys see a need to see more or less than the current default ?

[IgorK 111]

@Djtux You and @RapperT need to merge

[RapperT 336]

23 minutes ago, IgorK said:

@Djtux You and @RapperT need to merge

Djtux does great work.  I love his charts.  Im sure we could find some way to collaborate!

Edited September 19, 2017 by RapperT

[Kim 7,943]

14 minutes ago, Djtux said:

Yep, i can look into changing that. Probably more intuitive to have the current date as the reference date for the RV in %. I have to check if there isn't any issues doing that.

Yes i have not documented the 'start date' box, but it's actually the start date of the range of earning dates considered. For the moment, i have the EOD data but i don't have the full history of earning dates to go that far.

The idea would be : by default, the start date is 2 years before today's date (to have roughly 8 earning cycles), but in the future it could be changed to 3 years for example.

 

As for why the chart always shows few days before the current date, i personally like to see a bit of context, and that number of days shown before the current date is not configurable. Do you guys see a need to see more or less than the current default ?

Sorry, I misread the "start date".

Maybe make the start of the chart configurable? I agree that seeing few days prior the current date might be useful, but many times it would help to see what is the downsize from the current levels. You can obviously calculate it manually, but..

[jroback 5]

In a different post, I mentioned a similar rule of thumb to what jbthree is suggesting (I think), which would be to take the percentage decline in RV from T(x) to T(0) = (Tx-T0)/Tx and divide by the interval in days (x) to get percentage decline per day.  The rough rule of thumb would be that an RV decline of 1%/day could be covered by a short strangle.  One possible way to depict this would be to include a 1%/day trend line on the RV chart that starts at the black average T0 point and extends leftward to the end of the chart. This gives a visual reference to compare each of the individual RV curves against

[Djtux 548]

@IgorK @RapperT Thanks for the support guys. We will see what can be done. The important thing for me is to build something that is sustainable. If there are some analytics (like the RV charts) that are critical to one's trading, then it's important to have it sustainable.

 

4 minutes ago, Kim said:

Sorry, I misread the "start date".

Maybe make the start of the chart configurable? I agree that seeing few days prior the current date might be useful, but many times it would help to see what is the downsize from the current levels. You can obviously calculate it manually, but..

Are you talking about the "RV in % chart" specifically or in general ? For that chart, i could show few days before the current date (like now), but set the reference date at the current date. That way, it's easier to see the downside risk in % from the latest date.

I'm trying to keep the number of options at the minimum, and only add it if it's really useful. Having a lower number of combinations allows me to cache more aggressively the backtest.

2 minutes ago, jroback said:

In a different post, I mentioned a similar rule of thumb to what jbthree is suggesting (I think), which would be to take the percentage decline in RV from T(x) to T(0) = (Tx-T0)/Tx and divide by the interval in days (x) to get percentage decline per day.  The rough rule of thumb would be that an RV decline of 1%/day could be covered by a short strangle.  One possible way to depict this would be to include a 1%/day trend line on the RV chart that starts at the black average T0 point and extends leftward to the end of the chart. This gives a visual reference to compare each of the individual RV curves against

I have to think about that. The idea about the trend line with a 1%/day from the T0 point doesn't look too difficult to add, but i will need to test it to see if the chart is still readable.

 

I will come back to you when i have something to show, and we can see what could be tweaked.

[skydragon 51]

On 9/18/2017 at 11:10 AM, jbthree said:

I think it would be useful to see, on a percentage basis, declines between the current date and expiration. If RV tends to decline between 10% and 15% between now and expiration, but I know I can sell 3 cycles of strangles at 6% each for 18%, the trade would be one to consider now, regardless of how expensive the straddle is relative to the average.

But if I see that the % gained on the strangles will not cover the % change in RV in previous cycles now, I may see that, for example %gained on straddles > %decrease in RV in previous cycles, if I enter the trade 1 week before earnings release.

I would love to hear you think, as well as @RapperT,  @greenspan76 ,  @Yowster and others think.

I think I understand your line of thought, and I like it as a potential tool.   However...  Perhaps I'm not up to speed enough but, wouldn't the case be that you couldn't "know" that you would get 6% per cycle on a successive number of strangles.  Only on the current cycle would you be able to best predict (and not a sure thing) the strangle return.  As you look 2-3 weeks down the road the picture gets even less clear.  Or maybe I'm missing something?

[Kim 7,943]

11 minutes ago, Djtux said:

@IgorK @RapperT Thanks for the support guys. We will see what can be done. The important thing for me is to build something that is sustainable. If there are some analytics (like the RV charts) that are critical to one's trading, then it's important to have it sustainable.

 

Are you talking about the "RV in % chart" specifically or in general ? For that chart, i could show few days before the current date (like now), but set the reference date at the current date. That way, it's easier to see the downside risk in % from the latest date.

I think it's a good idea.

[jroback 5]

Here is an example, and yes it is hard to read.  Maybe someone has some ideas to improve readability?  

But if this trend line could be implemented properly, it has some interesting properties.  For example, I was moving it up and down (while retaining slope the same) and found that in the absence of gamma gains it is hard to close the IBM straddle profitably if entered at day -26, even if you offset losses by 1% a day using short strangles (that is the definition of the trend line: the amount of RV decline that can be offset with the short strangles).  That is, if you anchor the line on the left side at the day -26 values (the opposite of what I suggested above), the RV curves almost always stay under the trendline (I think only the orange and brown lines are exceptions to this).  On the other hand, if you anchor the trend line today (9/19/17) then its much easier to close for a profit in each of the last 8 cycles.  

I think this might be a similar idea to what RapperT is showing with his charts 

[jbthree 6]

I believe it should read 0% for each cycle on the earnings date. If I hover over each cycle on T-20, for example, I will see the the percentage the RV would have declined had we opened an ATM straddle on that particular date, and closed it on T-0. 

 

That way, by hover over each of the points on T-20 we can see, for example, that between T-20 and T-0, RV tends to decline by 25-30%. If I sold 3 iterations of straddles at 6-7% each, the stock would have to move some to break even.

 

But, maybe, if I hover over each of the points on T-14, I see that RV tends to decline by 10-12%. If I sold 2 iterations of straddles at 6-7% each, the stock would not not have to move. Thus, waiting would until T-14 would make more sense.

Edited September 19, 2017 by jbthree

[RapperT 336]

our return forecast basically does what you're describing, the chart will look a bit different depending when you open but the slope would be the same

 

[Yowster 9,174]

@jrobackI see what you are getting at, but I believe the math may be a bit off.  We typically collect about 6%-7% on the short strangle sales, so that 1% short strangle offset per day is a bit low.  I think its typically in the 1.2% - 1.4% per day range (remember that weekdays are ignored in this chart) and that difference adds up over multiple weeks.

 

 Also, when looking at things this far in advance (more than 20 days) treating the RV decline as a straight line may not be totally accurate.  Looking at the IBM chart in your post it took around 10 days for the average RV to drop 1% from T-26 to T-16, but then in those last 16 days the average only went down by about 0.6%.  So, for me, it appears that I may have entered my IBM trade about a week early as the rate of RV decline is higher during this past week that it is in the weeks closer to earnings day.

Edited September 19, 2017 by Yowster

[jbthree 6]

 

I believe it should read 0% for each cycle on the earnings date. If I hover over each cycle on T-20, for example, I will see the the percentage the RV would have declined had we opened an ATM straddle on that particular date, and closed it on T-0. 

 

That way, by hover over each of the points on T-20 we can see, for example, that between T-20 and T-0, RV tends to decline by 25-30%. If I sold 3 iterations of straddles at 6-7% each, the stock would have to move some to break even.

 

But, maybe, if I hover over each of the points on T-14, I see that RV tends to decline by 10-12%. If I sold 2 iterations of straddles at 6-7% each, the stock would not not have to move. Thus, waiting would until T-14 would make more sense.

 

If I wanted to do this now, I would hover over the date I am interested in, and get the RV (RV1) and then look at the RV at T-0 (or at final strangle expiration) (RV2), and then make the following calculation (RV1-RV2)/RV1 to get the percentage decrease from that date.

I would like to say I do this for each cycle, but I usually just make this calculation for the average. Each cycle would be more helpful. Moreover, as I mention above, I would like to quickly scan each of the dates to see when it may make sense to enter, or to realize that I missed my chance this earnings season.

Edited September 19, 2017 by jbthree
added quote for clarity.

[jroback 5]

3 minutes ago, Yowster said:

@jrobackI see what you are getting at, but I believe the math may be a bit off.  We typically collect about 6%-7% on the short strangle sales, so that 1% short strangle offset per day is a bit low.  I think its typically in the 1.2% - 1.4% per day range (remember that weekdays are ignored in this chart) and that difference adds up over multiple weeks.

 

 Also, when looking at things this far in advance (more than 20 days) treating the RV decline as a straight line may not be totally accurate.  Looking at the IBM chart in your post it took around 10 days for the average RV to drop 1% from T-26 to T-16, but then in those last 16 days the average only went down by about 0.6%.  So, for me, it appears that I may have entered my IBM trade about a week early as the rate of RV decline is higher during this past week that it is in the weeks closer to earnings day.

I agree with both points.  On the first one, I used 1%/day as a convenient estimate, although it might be low as you note.  Its also important to consider when the shorts expire relative to the earnings date because the premium collected prior to expiration will have to be "stretched' over the remaining couple days when you won't collect premium.

Your second point is what led me to this line of thinking. I was wondering why the IBM trade had declined so significantly, and then when I started playing around with the 1% trend line I noted that this behavior is actually quite typical when the IBM straddle is entered more than 20 days before earnings.  

Really interesting stuff. . . 

[Yowster 9,174]

1 minute ago, jbthree said:

If I wanted to do this now, I would hover over the date I am interested in, and get the RV (RV1) and then look at the RV at T-0 (or at final strangle expiration) (RV2), and then make the following calculation (RV1-RV2)/RV1 to get the percentage decrease from that date.

I would like to say I do this for each cycle, but I usually just make this calculation for the average. Each cycle would be more helpful. Moreover, as I mention above, I would like to quickly scan each of the dates to see when it may make sense to enter, or to realize that I missed my chance this earnings season.

Yes, that is the most important to me.  I want to see that adjusted RV when factoring in short strangle sale(s) to look good when compared to prior cycles.  The biggest gains come from gamma so this gives the stock more days to move.   Turns out I probably entered by IBM trade a bit early, but if the stock happening to move during the past week then it would have paid off  - but my adjusted RV after multiple sales should get me in good shape when compared to prior cycles so I'm not panicking at this point.