volatilityhq.com Official Thread

[JoeA 9]

Given the SO methodology, is it important to consider the RV of a double-calendar? I don't know if your site provides RVs for double calendars, but reading through a lot of the past trade discussions, it seems like Kim puts a lot of value on pricing a double, as opposed to a single, calendar. Thank you.

[Djtux 548]

3 minutes ago, JosephAnderson said:

Given the SO methodology, is it important to consider the RV of a double-calendar? I don't know if your site provides RVs for double calendars, but reading through a lot of the past trade discussions, it seems like Kim puts a lot of value on pricing a double, as opposed to a single, calendar. Thank you.

Yes you can select just calls, just puts or double calendar to compute RV for calendars. It's actually better to be able to see the exact RV for puts or calls separately to know the exact target price.

Here is a relevant discussion about that if you have access to SO forums : https://steadyoptions.com/forums/forum/topic/4289-pre-earnings-double-calendar-question/?do=findComment&comment=92673

 

[JoeA 9]

On 2/14/2018 at 10:24 AM, Djtux said:

Yes you can select just calls, just puts or double calendar to compute RV for calendars. It's actually better to be able to see the exact RV for puts or calls separately to know the exact target price.

Here is a relevant discussion about that if you have access to SO forums : https://steadyoptions.com/forums/forum/topic/4289-pre-earnings-double-calendar-question/?do=findComment&comment=92673

 

In previous discussions by Kim, I thought a double calendar was two different strikes near each other, e.g., 100 and 105 for IBM. Each could be either a put or a call calendar depending on how expensive they are. But your answer implies a double calendar is both puts and calls at the same strike. Am I misunderstanding something? Thanks. 

[Djtux 548]

19 minutes ago, JosephAnderson said:

In previous discussions by Kim, I thought a double calendar was two different strikes near each other, e.g., 100 and 105 for IBM. Each could be either a put or a call calendar depending on how expensive they are. But your answer implies a double calendar is both puts and calls at the same strike. Am I misunderstanding something? Thanks. 

You are right, for consistency with the RV charts from https://steadyoptions.com/forums/forum/topic/1393-tools-to-analyze-the-earnings-trades/, i also define double calendar as 1 call calendar and 1 put calendar at the ATM strike.

I think that whole discussion is confusing because now that the RV can be computed for puts or calls directly, that gives you the target RV directly for either a put calendar or a call calendar. And you are right that usually with SO we enter the ATM calendar (put or call) at 1 strike for half position and another calendar at ATM if the stock moves.

I'm not sure if that helps or not.

 

[Yowster 9,174]

2 minutes ago, Djtux said:

You are right, for consistency with the RV charts from https://steadyoptions.com/forums/forum/topic/1393-tools-to-analyze-the-earnings-trades/, i also define double calendar as 1 call calendar and 1 put calendar at the ATM strike.

I think that whole discussion is confusing because now that the RV can be computed for puts or calls directly, that gives you the target RV directly for either a put calendar or a call calendar. And you are right that usually with SO we enter the ATM calendar (put or call) at 1 strike for half position and another calendar at ATM if the stock moves.

I'm not sure if that helps or not.

 

For RV purposes on charts, the double calendar was always the call calendar and the put calendar at the same strike (I think this was done because sometimes puts and calls can trade at different prices and this was a way to factor in the difference).   Practically, we open whatever calendar is cheaper and will often open another calendar at a different strike if the stock price moves.   Summarizing, the double calendar RV in the charts is for the same strike and is used to indicate when the calendar price is trading at a good level based on prior cycles.   The double calendar in actual trades is calendars at different strikes (either centered around ATM when you open the trade or opening one ATM and opening another one later as the stock price moves).

[JoeA 9]

Thank you, djtux and Yowster. The matter is much clearer to me now. 

[Djtux 548]

Fyi, i've added the return matrix menu under a new beta category.

I'm still testing that and implementing some improvements. Performance might be slower for the return matrix as well.

[Djtux 548]

On 8/22/2017 at 3:51 PM, 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.

I believe the new return matrix would show that, except it's showing the return of a straddle entered at T-X and exited at T-Y.

But that's only for a straddle long only position, doesn't take into account any strangle selling like the hedge straddle strategy done with SO.

For example : https://www.volatilityhq.com/backtester/return_matrix/?symbol=QCOM&start_date=2017-02-17&strategy_type=Straddle&submit=Run+backtest

[Djtux 548]

On 7/13/2017 at 10:20 PM, cwelsh said:

maybe we read these charts differently?  I thought today was T-3?  That'd seem to match the line.

 

I also look at the rate of decline in RVs in making a determination.

 

Also, on these charts, I obviously love them and the daily RV calculations are useful.  But also what would be really nice is to see the change in RV carried forward.  This would end up with a lot more lines, but it'd go like (example only):

 

1.  T-10, Stock ABC is at 100, and the ATM straddle has an RV of 10%.  What is the value of THAT straddle at T-9, T-8, T-7, etc;

2.  Then we do the same at T-9.  If the price has moved to 95, we now have two lines, one starting at T-10 for the 100 straddle and one starting at T-9 for the 95 straddle;

3.  Repeat at T-8.  At T-8, what is the value of the 100 straddle, the 95 straddle, and let's say the price went back to 97.5, so we now have a 3rd line for the 97.5 straddle.

4. Etc.

This would probably really help pin down the "ideal" time to enter a trade.  What the current charts do is help us evaluate if entering a trade is "smart" (e.g. getting a good current value), but, as we've seen, we could be getting a great historical RV value, but would have been better off entering two or three days earlier or later.  The current RV charts won't show that.

 

In fact, the current charts almost always show RV declining every day you get closer (which of course makes sense with theta decay).  If these charts held true, then this strategy wouldn't work.

 

I've started to add the 'return matrix' that shows the median return of straddle position assuming it is entered at T-x and exited at T-Y, but i could modify the matrix to show the RV evolution each day. For example, you would see the RV of straddle entered at T-10, and follow the RV at T-9, T-8, until T-0 to see how it evolves.

A screenshot of a prototype i have on my dev environment (not available on the production website) is attached below. I chose BUD because we just traded this.

Let me know if that's close to what you have in mind.

 

Edit: i've also added the median return matrix for comparison purpose as well, which is on the production website right now. Also Friday 16th was T-7.

Edit 2: the RV calculated below is probably incorrect as i just realized that i was dividing the straddle price at T-Y by the stock price at T-X, but it's only on my machine and just to illustrate the idea.

 

Edited February 17, 2018 by Djtux

[Spread 21]

Hi @Djtux, this might be a silly newbie question. But I have a simple question about the RV calculation.

I'm putting my own worksheet together in excel using RTD data from TOS. So i would like to know the way RV is calculated. Is this "the straddle price / the strike" or "the straddle price / current price underlying". Thanks in advance.

Edited February 19, 2018 by Hielke

[Djtux 548]

1 hour ago, Hielke said:

Hi @Djtux, this might be a silly newbie question. But I have a simple question about the RV calculation.

I'm putting my own worksheet together in excel using RTD data from TOS. So i would like to know the way RV is calculated. Is this "the straddle price / the strike" or "the straddle price / current price underlying". Thanks in advance.

Relative value is usually define on SO forum as the strategy price / stock price. So for straddle, it would be straddle price / stock price.

[Djtux 548]

I've made some changes to the website :

A new flag to the Heat-map Return Matrix to show/hide the individual cycles :

I've also added a new scanner under the beta menu that works in pair with the Heatmap Return Matrix :

 

The beta features are ready to be used to get feedbacks while i'm continuing the development and bug fixes.

 

[Antoxa 14]

On 2018-02-17 at 8:41 AM, Djtux said:

Does this include gamma gains?

It feels like you take ATM Straddle at Entry and ATM Straddle at Exit and compare them. Am I wrong?

[Djtux 548]

5 minutes ago, Antoxa said:

Does this include gamma gains?

It feels like you take ATM Straddle at Entry and ATM Straddle at Exit and compare them. Am I wrong?

No if you go to the website, you can put the mouse in each cell of the heatmap matrix to see the 4 returns used to compute the median for a given entry T-X and exit T-Y.

Then for each earning cycle, you have 1 heatmap as well that details the return for each ATM straddle entered at T-X and exited at T-Y.

Once the ATM straddle is entered at a given T-X, the strike is not reset and you calculate the price each day to get the exit price.

So gamma is included.

What is not included are any adjustment done if the stock moves, and there are not strangle sold short to replicate the hedged straddle.

[StuC 5]

13 hours ago, Djtux said:

I've made some changes to the website :

A new flag to the Heat-map Return Matrix to show/hide the individual cycles :

I've also added a new scanner under the beta menu that works in pair with the Heatmap Return Matrix :

 

The beta features are ready to be used to get feedbacks while i'm continuing the development and bug fixes.

 

Very nice additions.  Thank you!

[Djtux 548]

Made some changes to display a new green line that shows the average stock move (it's like the price effect : abs (price close after earning / price close before earning - 1.0) ).

You also have the max move given even if the horizontal line is outside of the range so that the RV lines are not compressed.

That can be enabled or disabled with this flag (it is enabled by default) :

You will also notice that most flags are collapsed under an 'advanced accordion'. I though it would be easier not to have to scoll too much to see the RV charts. I'm open to change that if you don't like that.

[NikTam 163]

On 2/21/2018 at 8:27 PM, Djtux said:

I've made some changes to the website :

A new flag to the Heat-map Return Matrix to show/hide the individual cycles :

I've also added a new scanner under the beta menu that works in pair with the Heatmap Return Matrix :

 

The beta features are ready to be used to get feedbacks while i'm continuing the development and bug fixes.

 

Interesting!  I will be taking a close look at this....

[Djtux 548]

Just a quick update : I've done improvements in the code to support symbol changes like PCLN that is now under BKNG.

At this moment i have to enter manually in the database the mapping change, so if you are interested in a company that changed symbol, you can just point that out to me (either here, or PM, or by email to support(at)volatilityhq.com ).

I've also mapped BBRY to BB (Blackberry).

[Djtux 548]

Just a note : the status for the earning date #1 is now showing all the next earning dates are 'confirmed' and i'm fixing that right now.

Sorry for the inconvenience.

[Djtux 548]

1 hour ago, Djtux said:

Just a note : the status for the earning date #1 is now showing all the next earning dates are 'confirmed' and i'm fixing that right now.

Sorry for the inconvenience.

This is fixed now.

[Djtux 548]

Some of you might have issues with the website as it is hosted on AWS US East (North Virginia).

According to AWS status page https://status.aws.amazon.com/, it is supposed to be fixed, but i believe it's not totally the case.

Quote

10:24 AM PST Between 6:23 AM and 6:32 AM PST and between 8:11 AM and 8:21 AM PST, customers in the US-EAST-1 Region may have experienced Internet connectivity issues. The root cause of this issue was a loss of power in one of our network peering facilities. Our network is designed to be fully redundant with multiple independent peering facilities in every region. Some customers experienced elevated latency and packet loss while the network rerouted affected traffic to these unaffected network peering facilities. Some packet loss was also observed as we restored traffic to the affected network peering facility. The issue has been resolved and the service is operating normally.

Another link less 'technical', good old CNBC : https://www.cnbc.com/2018/03/02/amazon-cloud-networking-outage-affecting-atlassian-twilio-slack.html

Quote

Amazon scrambles to fix cloud networking issue affecting companies like Atlassian, Twilio

Amazon Web Services acknowledged a networking issue at about 10:30 a.m. Eastern time on Friday.
Atlassian, Slack, Twilio and other web services reported problems that appeared to be related to the cloud issue.

One of the first web services to go down was Amazon's own voice assistant, Alexa, as TechCrunch reported.

GitHub, MongoDB, NewVoiceMedia and Slack were among the companies that disclosed issues midday on Friday.

 

Edited March 2, 2018 by Djtux

[Djtux 548]

Made some changes to the heatmap return matrix :

Now there is only 3 strategy types (ATM straddle, or long call or long put).

For the long call or the long put, you can also select the delta :

Some remaining improvements not done yet around the heatmap return matrix and the return matrix scanner :

• add more combinations to the return matrix scanner : more deltas. more strategies, etc
• add support for additional strategies : maybe like calendars or iron condors (which increase the complexity a little bit to do it properly : have same width on both sides, compute the return properly on margin not credit receive)
• add support for post-earning strategies instead of pre-earning strategies like now

[PGAmbrosio 53]

@Djtux just wanted to ask, around what time your scanner data gets updated with the latest EOD data?  Also, do you have any plans for a favicon =D?  Thanks in advance and great stuff on volatilityHQ.

[Djtux 548]

8 minutes ago, PGAmbrosio said:

@Djtux just wanted to ask, around what time your scanner data gets updated with the latest EOD data?

Depends on when the data provider publish their files for the day, but i would say around 6 or 7 pm Eastern time (just a ballpark number). I've had problems in the past because of the background workers (Python celery workers for the technically minded) kept dying without being restarted and for a long time i had to babysit that. I recently found the root cause that was difficult to find and i believe that fixed the issue, so each day the data import and update of the scanners should be more consistent and reliable.

16 minutes ago, PGAmbrosio said:

Also, do you have any plans for a favicon =D?

Ah yes forgot about that, i can put one, but it's probably going to be kind of ugly ;).

[Djtux 548]

@PGAmbrosio Actually the scanner could be updated as soon at 5:15 pm Eastern. The scanner is up to date right now.

Btw, if you are not sure what is has been used in the scanner, the column "EOD Data Date" indicates what was the latest data being used, so if it is today then the data for today has been imported.

 

Also i did a quick&dirty favicon, might need to force the refresh on Google Chrome (CTRL+F5) to "enjoy" the art.

[PGAmbrosio 53]

@Djtux I was aware of that last column, which is why I asked.  The scanner is a valuable tool for me and I like use it quite often.  Also, GREAT work on the art, really, it is much better than the blank page icon!

[Djtux 548]

I made some changes to the scanner.

• The filters for the scanner now are done server side, so you need to submit the form. The reason i did that was to reduce the number of data being sent from the server to your browser.
• I added a new column : S.M. for the bid-ask spread of the ATM options in % of the stock price. So it's the average of (ask call-bid call) and (ask put -bid put). And then i divide that average by the stock price. The strikes are selected to be the ATM (so the closest to the stock price). The expiry is the first monthly expiration after the earning date. That should give another way to filter tickers that have wide spreads.

[Djtux 548]

Seems the data provider is late to publish the EOD data, not sure if it's related to the DST or something else. I sent a message.

[MichiganWater 11]

Hi Djtux,

I just tried running the scanner with all fields empty and "no" and "no" for the Only Confirmed and Only Weekly fields, but it only pulls 62 entries.  It didn't pull, for example, RHT.

Thanks

[Djtux 548]

26 minutes ago, MichiganWater said:

Hi Djtux,

I just tried running the scanner with all fields empty and "no" and "no" for the Only Confirmed and Only Weekly fields, but it only pulls 62 entries.  It didn't pull, for example, RHT.

Thanks

I see, there is an issue with the scanner job that didn't produce everything. I will do a temporary fix today.

[Djtux 548]

19 minutes ago, Djtux said:

I see, there is an issue with the scanner job that didn't produce everything. I will do a temporary fix today.

@MichiganWater I just did a temporary fix, If removed all scanner filters i see "Showing 1 to 50 of 1,409 entries". The root cause is that Amazon AWS service i'm using as a cache is a little limited in size for my needs (so i'm caching more and more stuff), so i will do deeper work in the coming weeks to improve that. Sorry for the inconvenience.

[MichiganWater 11]

4 hours ago, Djtux said:

@MichiganWater I just did a temporary fix, If removed all scanner filters i see "Showing 1 to 50 of 1,409 entries". The root cause is that Amazon AWS service i'm using as a cache is a little limited in size for my needs (so i'm caching more and more stuff), so i will do deeper work in the coming weeks to improve that. Sorry for the inconvenience.

Thank you.  I just tested it and it's working great now.

[Djtux 548]

On 3/13/2018 at 9:15 PM, Djtux said:

I just did a temporary fix, If removed all scanner filters i see "Showing 1 to 50 of 1,409 entries". The root cause is that Amazon AWS service i'm using as a cache is a little limited in size for my needs (so i'm caching more and more stuff), so i will do deeper work in the coming weeks to improve that. Sorry for the inconvenience.

The issue with the scanner reappeared, i'm trying to do another temporary fix, i will need to weekend to do the deeper work.

[Djtux 548]

58 minutes ago, Djtux said:

The issue with the scanner reappeared, i'm trying to do another temporary fix, i will need to weekend to do the deeper work.

Need to take down the website to upgrade the caching servers to have more memory.

[Djtux 548]

55 minutes ago, Djtux said:

Need to take down the website to upgrade the caching servers to have more memory.

Upgrade done, that should fix the scanner issues.

[NikTam 163]

Nice! 

[Djtux 548]

Made some improvements to the heatmap return scanner :

In addition to the 40 delta call that was already there, i added ATM straddle, 60 delta call, 50 delta call and 40 delta put.

[Sirion 93]

@Djtux Suggestion for a feature adjustment, though it's not too much trouble to manually do this:

On your Earnings Feed page, can you provide an option to sort by Earnings Date instead of confirmation date?

Secondly, a clarification question: Is the Source column the "total number of sources that have confirmed" or an identifier to which source most recently confirmed?

[Djtux 548]

Just now, Sirion said:

Suggestion for a feature adjustment, though it's not too much trouble to manually do this:

I always appreciate suggestion, i might see what could be useful for some people sometimes.

1 minute ago, Sirion said:

On your Earnings Feed page, can you provide an option to sort by Earnings Date instead of confirmation date?

Can do that, shouldn't be too difficult.

2 minutes ago, Sirion said:

Secondly, a clarification question: Is the Source column the "total number of sources that have confirmed" or an identifier to which source most recently confirmed?

For now it just indicates which source has confirmed.

 

One thing that would make the earning date feed more useful is that it should only show a ticker the first time one source confirms, and after that it should not repeat for other sources, or else you get confused.

[Sirion 93]

Just now, Djtux said:

I always appreciate suggestion, i might see what could be useful for some people sometimes.

Can do that, shouldn't be too difficult.

For now it just indicates which source has confirmed.

 

One thing that would make the earning date feed more useful is that it should only show a ticker the first time one source confirms, and after that it should not repeat for other sources, or else you get confused.

I'd think the most useful iteration would be to keep a floating tracker of which sources have confirmed. If #1 previously confirmed, then #4, refresh the "ticker" and have #1, #4 under source column. Maybe bold #1 to indicate it's the most recent and update the timer for it?  Personally, I feel more confident using multiple confirmations - I know you show this on the Straddle page, but the return matrix doesn't have that feature yet (maybe worth copying that code over if it isn't resource intensive?)

Does it ever happen that they confirm different dates? (ie, somebody messed up)

(also, I have no idea which sources are which number, though I'm sure you're consistent throughout the site  )  

 

Thanks for the quick response!

[Yowster 9,174]

@Djtux A minor tweak suggestion.   When going through a bunch of candidates and pulling up straddle RV charts, I sometimes get fooled thinking that rate of RV decline is less than what it is because the y-axis increments are 2% instead of the normal 1%.    What do you think of always using 1% y-axis increments so when looking at rate of RV decline it will always be the same scale?

[Djtux 548]

3 minutes ago, Yowster said:

@Djtux A minor tweak suggestion.   When going through a bunch of candidates and pulling up straddle RV charts, I sometimes get fooled thinking that rate of RV decline is less than what it is because the y-axis increments are 2% instead of the normal 1%.    What do you think of always using 1% y-axis increments so when looking at rate of RV decline it will always be the same scale?

I think it's a good point.

Let's take the SCHW example. You would get the same chart except that on the y-axis, you would get more horizontal line in the grid so that there is 1 horizontal line at each 1% increment ?

 

[Yowster 9,174]

@DjtuxYes, and same spacing between.  So, in your example the 4/6/8/10% would be 4/5/6/7/etc% with the same spacing.   The rate of RV decline would have a bigger downward slope than what it looks like now.   You get used to looking at charts with 1% increments as most stocks are that way, so when you get a chart with bigger increments sometimes the RV line appears flatter than what it really is.

[Djtux 548]

4 minutes ago, Yowster said:

@DjtuxYes, and same spacing between.  So, in your example the 4/6/8/10% would be 4/5/6/7/etc% with the same spacing.   The rate of RV decline would have a bigger downward slope than what it looks like now.   You get used to looking at charts with 1% increments as most stocks are that way, so when you get a chart with bigger increments sometimes the RV line appears flatter than what it really is.

I think i will need to think a little bit as if i represent a 1% RV increment in y-axis using the same spacing, then automatically i constraint the min/max range of RV i can represent.

So for some stocks, either the RV lines might be compressed together, or for others the RV lines might be outside the min/max range.

I'm not sure if you see what i mean.

I will have to do some tests to find a compromise that works for most stocks.

[Yowster 9,174]

@DjtuxI see what you mean, its often one outlier cycle that requires the bigger axis - in this case its the one cycles where the closest expiration to earnings was 4 weeks out.

[cwelsh 1,549]

On 3/30/2018 at 9:41 AM, Yowster said:

@DjtuxI see what you mean, its often one outlier cycle that requires the bigger axis - in this case its the one cycles where the closest expiration to earnings was 4 weeks out.

I've noticed the axis thing a few times too - in those cases I've just replotted on an excel spreadsheet of my own, but maybe a future feature (as if you're not way above and overboard adding things all of the time anyway) would be a way to either (a) adjust the y axis or (b) export the graph/data straight to excel so we could manipulate. 

 

(not complaining, I know you're constrained by displaying on a screen, so if the intervals of RV are bigger, particularly due to an outlier, the y axis has to be adjusted - not sure what the better solution would be - maybe give us the opportunity to delete an outlier and change the axis??)

[Djtux 548]

Just a quick message to let you know i'm aware that 1 of the web server is in a degraded state and i'm currently redeploying the web servers, so it should be back to normal soon.

[Djtux 548]

On 3/31/2018 at 11:49 PM, cwelsh said:

I've noticed the axis thing a few times too - in those cases I've just replotted on an excel spreadsheet of my own, but maybe a future feature (as if you're not way above and overboard adding things all of the time anyway) would be a way to either (a) adjust the y axis or (b) export the graph/data straight to excel so we could manipulate. 

 

(not complaining, I know you're constrained by displaying on a screen, so if the intervals of RV are bigger, particularly due to an outlier, the y axis has to be adjusted - not sure what the better solution would be - maybe give us the opportunity to delete an outlier and change the axis??)

I always appreciate the feedbacks and improvement ideas. It doesn't mean they get implemented right away, but i store everything in a 'todo list'.

Yes it takes time to handle the production side (keep the data ingestion smooth, and the servers running), but over time new features will be added.

I'm not sure exactly how to address the y-axis and the outliers issue yet, i have to do some tests.

[Djtux 548]

I'm aware of an issue with earning dates retrieval that causes some issues with BKNG/PCLN, i don't have a timeline for a fix yet.

[Djtux 548]

11 hours ago, Djtux said:

I'm aware of an issue with earning dates retrieval that causes some issues with BKNG/PCLN, i don't have a timeline for a fix yet.

Should be fixed now in production.

You might have noticed some generic error message for some symbols today, but that should also be fixed.

I had some issues with the source #4 for the earning dates.