volatilityhq.com Official Thread

[Djtux]

On 10/20/2017 at 7:48 PM, Djtux said:

The D and CD columns that is in the scanner shows 3 and 4 respectively because it is calculated with the 'today' date UTC time.

Decay BD (5.82%/6.00%-1)/4 = -0.75%

 

3 hours ago, Christof+ said:

@Djtux, I am thankful for your useful comment on art-of.trading, here a constructive remark on the decay calculation above:

If your goal is to convey an idea about the actual daily percentage decay of a position held in the straddle (f.e. for the short theta you would need on any day to offset the effect), a simple average will not do the trick. Returns compound. Since the basis of the percentage calculation changes every day, they are not additive. For a correct calculation you would have to use the geometric mean (either by using 1 + total_decay = (1 + daily_decay)^ nbr_days, or by simply converting to continuous/log returns which are additive). I would argue that for very small decays this is an academic exercise with not much practical value. Unfortunately, RV decays can be quite large. For example, for an RV losing 50% over 20 days, the daily decay you as a straddle holder would be exposed to is 3.41%, while the simple average is just 2.50%.

A simple daily average will always underestimate the true value (I have no access to your site but I would expect your values to be generally smaller than mine). Additionally, the error grows exponentially with an increase in total decay (rendering comparative statements like 'RV of YY decays twice as fast as the one of ZZ' incorrect).

 

First of all, this is the official thread for my web service and i would appreciate if you don't put links back to your own service in this specific thread : that includes the link in the post and the signature.

For the decay, yes i'm aware of the different conventions (simple, compounded, exponential/continuous). To be technically correct, depending on how you count the periods, the simple convention could be lower or higher than the compounded version. For example, if you count in year fraction and it is < 1.0, then the simple compounding gives a higher value. This is why for example that the simple convention is used in shorter-term money market instruments or car loan.

Here is the goal of the decay is to answer those questions :

• Is the RV not declining too fast so that a given ticker is a candidate for a straddle ?
• Can i get enough credit (say 6% of the straddle) with a strangle to compensate the RV decline ?

The decay formula above are equivalent to a crude linear regression (degenerate as it only takes 2 points) with x being the days (calendar or business) and y being the RV in % (1.0 being the reference date).

Or equivalently it means you draw a straight line in the 'RV in %' chart and the decay is the slope.

That makes it easy to compute an approximate decay for the next 5 days and compare with the credit you would get by selling a strangle.

Or another way to use that decay column is to set some filters to see the ones where there is no way the strangle credit would cover the decline so you don't lose time analyzing tickers that declines too fast.

To illustrate, let's see on the hypothetical example you give :

Quote

For example, for an RV losing 50% over 20 days, the daily decay you as a straddle holder would be exposed to is 3.41%, while the simple average is just 2.50%

So at T-20 the RV in % is 100%. At T=0, the RV in % is 50%.

2.5% is the slope of the line (blue line in the screenshot below) that goes between the 2 points.

3.41% is the orange line : compounded.

I still think that my modelling is still easier to understand and answer the 2 questions above (RV declining too fast and getting enough credit for the strangle).

I could also improve the linear regression to take more points also instead of just the points at T-20 and T-0.

If anyone else has an opinion on this, i would be interested to hear it.

 

[MichiganWater]

3 hours ago, Djtux said:

If anyone else has an opinion on this, i would be interested to hear it.

 

For me, I'm happy with it the way it is.  For the straddles I'm interested in those with very little decay, and the difference between the two methods will not impact my use of the scanner to screen candidates.

On the flip side, I'm considering a strategy for the stocks with fast decay that sells the week-of straddle (or strangle) and purchases a wider strangle the next week.  Sort of like an Iron Condor Calendar, with the purpose of making the wider strangle further out in time to mitigate the IV increase due to earnings.  Even with that, though, my screening will be looking at Decay's relative to other Decay's, so it won't matter if it's simple vs compounded.

[Yowster]

12 minutes ago, MichiganWater said:

For me, I'm happy with it the way it is.  For the straddles I'm interested in those with very little decay, and the difference between the two methods will not impact my use of the scanner to screen candidates.

On the flip side, I'm considering a strategy for the stocks with fast decay that sells the week-of straddle (or strangle) and purchases a wider strangle the next week.  Sort of like an Iron Condor Calendar, with the purpose of making the wider strangle further out in time to mitigate the IV increase due to earnings.  Even with that, though, my screening will be looking at Decay's relative to other Decay's, so it won't matter if it's simple vs compounded.

@MichiganWater One comment to be aware of when looking for straddles with very little decay.   Typically, stocks whose straddles show little decay are those for higher IV stocks - and the downside for straddles is that higher IV stocks have slower gamma gains when the stock price moves. 

 

Also, for your proposed trade for the faster decaying stocks, what short to long ratios are you looking at?    This is kind of a totally different trade, IMO, so ratios may be quite different and the stocks you look for may be those that show little IV run-up into earnings.   Would be interesting to see some backtests of this strategy.

[MichiganWater]

10 hours ago, Yowster said:

@MichiganWater One comment to be aware of when looking for straddles with very little decay.   Typically, stocks whose straddles show little decay are those for higher IV stocks - and the downside for straddles is that higher IV stocks have slower gamma gains when the stock price moves. 

 

Also, for your proposed trade for the faster decaying stocks, what short to long ratios are you looking at?    This is kind of a totally different trade, IMO, so ratios may be quite different and the stocks you look for may be those that show little IV run-up into earnings.   Would be interesting to see some backtests of this strategy.

@Yowster That's a good point, and it's something I look at when I'm considering straddles, typically through seeing how much the price of straddles increases as the strikes move away from ATM.  I haven't figured out a way to model that, so for now it's purely a subjective evaluation.

I don't want to extend the discussion of the proposed strategy in this thread, since it's Djtux's dedicated thread for his service, but since it's an application that makes use of his service I figure it's ok to just give my logic.  I haven't developed the method or backtested anything yet, but I'm organizing my thoughts as follows:

a) Straddles that are horrible for long positions because of decay should be good candidates (under some circumstances) for shorting, and I can find these with Djtux's Straddle tool.

b) I never naked short anything, so I'd use an Iron Butterfly setup to short the straddle.  I could also make these into an Iron Condor, but that requires the assumption that the decay data calculated by volatilityHQ for straddles applies to similar strangles.  That's an assumption I'm willing to make, but it's an assumption nonetheless.

c) Iron Butterflys/Condors are short vega, which is the wrong thing to be going into earnings, so to "fix" that I look to the long-vega nature of calendars.  That's why I choose a later expiry for the long strangle.  This is also another application of volatilityHQ, using the Calendar tool.  If the relevant calendar gains significantly, then great.  If not, then I wouldn't expect the setup to work well.

So, in summary, if the short straddle shows big decay and the long calendar shows a good increase, I'm guessing that that an "Iron Butterfly/Condor Calendar" that doesn't move a lot before earnings will be a winner, on average.

I haven't started messing with ratios yet, or anything like that yet.  I'm nowhere near an expert in options trading, so I don't think that _I_ want to open a thread to discuss this strategy, but if you (or anyone else) wants to investigate and thinks it's worth it, please feel free to start a thread!

Djtux, hopefully you're not opposed to me posting this content here, since it shows how to apply volatilityHQ using both the Straddles and Calendars tools, but if you would rather have it moved out of your service thread, then please do so.

[Djtux]

On 8/20/2017 at 1:00 PM, Djtux said:

"Volatilityhq is officially launching. We hope that you will support us to keep this going. The regular monthly price is $99.99/month. To say thank you, there is a discount code of 50% only available at launch. 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). The discount code is LAUNCH and must be redeemed by Oct 31st, 2017. In case of issues, email support(at)volatilityhq.com."

Just a reminder that today is Oct 31st, 2017 if you want to take advantage of the coupon. 

[Djtux]

1 minute ago, krisbee said:

I will answer here.

Yes it looks wierd, but i believe it's normal because that far out in advance, the selected expiry is not trading that far in advance.

For example, if i take the 2017-11-17 expiry, it only started trading on 09/18/2017 (i checked on ThinkOrSwim Thinkback) which is what you see in the orange line.

The RV in % can't work because we are too far in advance.

 

[Dave W]

 

@Djtux, The service looks interesting. I signed up to try it out. Getting some scenario-based written training or videos on how to use each chart and table would be very helpful. Ping me via email if you might want help with something like that.

One question -- are you concerned that using T-30 days to expiration charts might be allowing you to 'peek' into the future on your backtests? In other words, let's say earnings are not confirmed until 20 trading days, or roughly 1 month, prior to the earnings event. If you show a chart using T-30, you would be peeking into the future by 10 trading days to have knowledge of the right expiration to use to display in the chart. I'm guessing this is why Dustin only used T-10 days -- because the vast majority of tradable options have announced the earnings date by then.

Or do you have a source of data so that you would only show T-30 when you know that earnings were confirmed prior to that date? In a perfect world, you would only show T- x days, where X = the number of days prior to earnings when earnings was confirmed.

Hopefully the point I'm trying to make is clear; if not, let me know and I'll try to clarify.

[Djtux]

12 hours ago, Dave W said:

The service looks interesting. I signed up to try it out.

Thank you for subscribing.

12 hours ago, Dave W said:

Getting some scenario-based written training or videos on how to use each chart and table would be very helpful. Ping me via email if you might want help with something like that.

Yes i definitely should improve the help section to explain all the concepts, but there are also many discussion threads on the unofficial trade ideas of the SO forum that discuss potential tickers in relation to RV charts.

If you have specific questions, please send me a PM or an email, i would be happy to answer those.

12 hours ago, Dave W said:

One question -- are you concerned that using T-30 days to expiration charts might be allowing you to 'peek' into the future on your backtests? In other words, let's say earnings are not confirmed until 20 trading days, or roughly 1 month, prior to the earnings event. If you show a chart using T-30, you would be peeking into the future by 10 trading days to have knowledge of the right expiration to use to display in the chart. I'm guessing this is why Dustin only used T-10 days -- because the vast majority of tradable options have announced the earnings date by then.

Or do you have a source of data so that you would only show T-30 when you know that earnings were confirmed prior to that date? In a perfect world, you would only show T- x days, where X = the number of days prior to earnings when earnings was confirmed.

Hopefully the point I'm trying to make is clear; if not, let me know and I'll try to clarify.

I'm not sure i understand at 100% your question, but i will try to give an answer, you tell me if it's answering what you have in mind.

I don't have the data yet to know historical earning "announcement" date (meaning when an earning date was confirmed), that was an idea i was looking into when we did the ORCL trade in June to speculate on the earning date announcement.

I think that for most tickers, the earning cycle seems predictable, and by looking from T-30, you will see either in the RV or the IV if an earning date was earlier than predicted as you will see the IV jump up (say the earning is moved from week N to week N-1).

Also looking at just T-10 might not be far enough for some tickers if you want to enter a hedged straddle (straddle+strangle strategy).

So, at this moment, i'm less concerned by that issue compared to say, making sure the next earning date is confirmed or not.

 

[Dave W]

7 hours ago, Djtux said:

 

 

I'm not sure i understand at 100% your question, but i will try to give an answer, you tell me if it's answering what you have in mind.

I don't have the data yet to know historical earning "announcement" date (meaning when an earning date was confirmed), that was an idea i was looking into when we did the ORCL trade in June to speculate on the earning date announcement.

I think that for most tickers, the earning cycle seems predictable, and by looking from T-30, you will see either in the RV or the IV if an earning date was earlier than predicted as you will see the IV jump up (say the earning is moved from week N to week N-1).

Also looking at just T-10 might not be far enough for some tickers if you want to enter a hedged straddle (straddle+strangle strategy).

So, at this moment, i'm less concerned by that issue compared to say, making sure the next earning date is confirmed or not.

 

@Djtux, Yes, I think you captured the issue - I was referring to knowing the historical date when the announcement date was confirmed. As you said, the cycle is reasonably predictable.

 

But I think we still have cases with large equities where we don't know if an announcement will fall in week X or week Y. I realize you have other concerns (a number of which you listed earlier in the thread). That said, I'll just be a voice of caution on this topic. I'd prefer to have as little "unknowable" knowledge as possible when we backtest. Knowing the actual date of an earnings announcement 30 days prior to earnings when we wouldn't have really known the date until 20 days prior to earnings could allow someone to come up with misleading backtest results. Hopefully you'll have the data to handle this at some point in the future.

 

You'll have to indulge me a bit on this particular topic; I was a partner in a firm that evaluated backtests, and the process is fraught with danger without adding peering into the future into the mix.

I have no doubt the tools will be useful despite what I've brought up. Appreciate you making them available.

Edited November 11, 2017 by Dave W

[Djtux]

I'm aware of the issues for some tickers on the site. I'm not 100% sure yet of the root cause but i think it might be caused by some errors coming from one of the data provider.

I will update when i have more information.

Sorry for the inconvenience.

[Djtux]

On 11/18/2017 at 6:52 PM, Djtux said:

I'm aware of the issues for some tickers on the site. I'm not 100% sure yet of the root cause but i think it might be caused by some errors coming from one of the data provider.

I will update when i have more information.

Sorry for the inconvenience.

I did some fixes Saturday night even if i wasn't able to pinpoint exactly the source of the issue (seems it's coming from a 3rd party library). It seems something is causing the web servers to be 'corrupted' and restarting them worked since Saturday evening until tonight.

I have other potential fixes to deploy tonight to see if it helps or not.

[RapperT]

Frustrating.   We’ve had issues with every data source we’ve tried 

[Djtux]

On 11/20/2017 at 8:34 PM, Djtux said:

I did some fixes Saturday night even if i wasn't able to pinpoint exactly the source of the issue (seems it's coming from a 3rd party library). It seems something is causing the web servers to be 'corrupted' and restarting them worked since Saturday evening until tonight.

I have other potential fixes to deploy tonight to see if it helps or not.

I think the issue disappeared. I've been monitoring that this week and the fix seems to have fixed it.

I did some improvements as well : 

• the scanner now monitors about 1200 symbols instead of about 400. You can always manually type the symbols in the charts page if the symbol you want is not there as i support all US stocks with options.
• to help filter all those candidates, i added additional columns : market capitalization of company, where or not there are weekly options, and a 'daily average options total volume averaged over 90 calendar days' which should help filter out the most illiquid stocks.

[krisbee]

is chart correct? first time i'm seeing the blue dotted horizontal line with a slope.

[Djtux]

49 minutes ago, krisbee said:

is chart correct? first time i'm seeing the blue dotted horizontal line with a slope.

Seems like a bug in the 2017-12-06 cycle as you see no blue RV line, and you also see a "RV nan%". I will investigate what's happening

Also note that the volume is low for options on SIGM, the stock is currently at 6.10 and the strikes for Dec15 is 5 and 7.5.

I think that causes some issues in the RV chart calculation as it might have a hard time selecting the ATM strikes (the strikes are further than +/-10% of the stock close)

 

 

[Djtux]

Just an FYI, i need to do some stronger maintenance on the backend tonight so the website might be slower than usual.

[Djtux]

Also for those that consider subscribing, a reminder of this offer :

Quote

 

Link to the website : https://www.volatilityhq.com/payments/confirm/1/?coupon_code=SO5017Q4 (you need to create an account first).

For members of this forum, there is a 50% forever coupon is SO5017Q4 and must be redeemed by Dec, 31st 2017. 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).

 

 

[Djtux]

I have extended the 50% coupon for those of you that are still interested.

Quote

For members of this forum, there is a 50% forever coupon is SO5018Q1 and must be redeemed by March, 31st 2018. 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).

 

[krisbee]

Current scanner excel sheet data for WBA is

I'm not sure if it'll be readable, so showing S thru AA columns below

 

Straddle chart of WBA has different number in many places? 

What time the chart gets updated? (I believe u get the feed everyday at 5pm)

What time the excel sheet scanner gets updated?

I'd like to know when the chart and scanner data will be in sync.

[Djtux]

8 minutes ago, krisbee said:

Current scanner excel sheet data for WBA is

I'm not sure if it'll be readable, so showing S thru AA columns below

 

Straddle chart of WBA has different number in many places? 

What time the chart gets updated? (I believe u get the feed everyday at 5pm)

What time the excel sheet scanner gets updated?

I'd like to know when the chart and scanner data will be in sync.

Sorry about that.

Yes i had some issues yesterday with the background workers and there is just a tons of jobs to catch up. Once the catch up is done (probably will be resorbed during the weekend), it should be back to normal.

During normal times i would say somewhere around 6pm maybe 7pm, i have to recheck when it's back to normal that it's still the case and see what i can improve.

I could add a column in the scanner to specify the date used to compute the "current RV", that way you know exactly if it is up-to-date or not.

 

[krisbee]

I just check EBAY Straddle chart. I don't see the latest BLUE line point for yesterday. 

Any reason?

 

[Djtux]

6 minutes ago, krisbee said:

I just check EBAY Straddle chart. I don't see the latest BLUE line point for yesterday. 

Any reason?

Yep, something happened but i'm not sure what exactly, the background workers (which are running on Amazon ECS) were not working properly. The container are up but the underlying Celery process wasn't, which shouldn't happen so i'm a bit puzzled.

I've manually restarted the background workers so all the jobs from the last 24 hours are backlogged but should be resorbed going forward.

Sorry about the inconvenience. 

[mustafaoe]

@Djtux Could you highlight the difference between the site https://www.art-of.trading and your service https://www.volatilityhq.com?  

[Djtux]

4 hours ago, mustafaoe said:

@Djtux Could you highlight the difference between the site https://www.art-of.trading and your service https://www.volatilityhq.com?  

I don't want to comment specifically about another service but volatilityhq.com started in beta for free around June of 2017 and became a paying service in August 2017.

I received from the members numerous bug reports and new feature requests over all that time and still have plenty of new features to implement in my backlog.

In my opinion, a service like volatilityhq.com is essential to the strategies being done on SO to help determine stock candidates and entry price.

Keep in mind that if you make just 1 bad trade overpaying to enter the trade, you will probably lose more than the $50.

Also there is a lot of effort and cost behind providing such service : data, software development/maintenance, servers cost (hosted on Amazon AWS), devops/ops, and i don't even talk about the non-technical aspect (legal, accounting etc). Can a service like this be free long term and be sustainable ? I don't think so.

Anyway, i don't think i'm the right person to answer your question, so i can only invite you to signup for 1 month and form your own opinion or rely on the opinion of members that are current subscribers.

[mustafaoe]

6 minutes ago, Djtux said:

I don't want to comment specifically about another service but volatilityhq.com started in beta for free around June of 2017 and became a paying service in August 2017.

I received from the members numerous bug reports and new feature requests over all that time and still have plenty of new features to implement in my backlog.

In my opinion, a service like volatilityhq.com is essential to the strategies being done on SO to help determine stock candidates and entry price.

Keep in mind that if you make just 1 bad trade overpaying to enter the trade, you will probably lose more than the $50.

Also there is a lot of effort and cost behind providing such service : data, software development/maintenance, servers cost (hosted on Amazon AWS), devops/ops, and i don't even talk about the non-technical aspect (legal, accounting etc). Can a service like this be free long term and be sustainable ? I don't think so.

Anyway, i don't think i'm the right person to answer your question, so i can only invite you to signup for 1 month and form your own opinion or rely on the opinion of members that are current subscribers.

Thanks for your reply. I will test it for 1 month

[zxcv64]

 

@djtux

I'm really enjoying my subscription to your charting site. I have a request/suggestion - in the same way that you have a 'column visibility' option, would it be possible to have a 'row visibility' or stock visibility option? Currently, there are dozens/hundreds of stocks listed in the scanner, and I only wish to see the ones which have a stocks above a certain price, or ones which have weeklies, etc. If there was a way to 'block' some stocks (based on an individuals preference), so that this stock would not appear in the scanner list then that would really help save wading through many pages. A simple check box against each stock symbol would do the trick. (Just realised it a bit more tricky than this as you would want to show another page where every 'blocked' stock is listed in case the user wants to unblock it at some later stage.)

Many thanks.

 

[Djtux]

1 minute ago, zxcv64 said:

 

@djtux

I'm really enjoying my subscription to your charting site. I have a request/suggestion - in the same way that you have a 'column visibility' option, would it be possible to have a 'row visibility' or stock visibility option? Currently, there are dozens/hundreds of stocks listed in the scanner, and I only wish to see the ones which have a stocks above a certain price, or ones which have weeklies, etc. If there was a way to 'block' some stocks (based on an individuals preference), so that this stock would not appear in the scanner list then that would really help save wading through many pages. A simple check box against each stock symbol would do the trick. (Just realised it a bit more tricky than this as you would want to show another page where every 'blocked' stock is listed in case the user wants to unblock it at some later stage.)

Many thanks.

 

I can see if it's possible to implement that with the 3rd-party table library that i use, but it's getting quite complex to add many filters.

In the meantime, did you know you could download an excel or csv file, import that in excel or google sheets and filter there ?

[zxcv64]

Just now, Djtux said:

In the meantime, did you know you could download an excel or csv file, import that in excel or google sheets and filter there ?

Thanks for the quick response. Yes, I have used the download function into EXCEL, but I thought if there was a function like the one I mentioned, then it would only need to be set once, and then in every future cycle any un-wanted stocks would then not appear on my list of stocks.

Just something to think about maybe at some point in the future, if possible.

 

[StuC]

9 minutes ago, Djtux said:

I can see if it's possible to implement that with the 3rd-party table library that i use, but it's getting quite complex to add many filters.

In the meantime, did you know you could download an excel or csv file, import that in excel or google sheets and filter there ?

This is what I do.  It's really simple and fast.  I can download into excel and have precisely what I want in 5 minutes by sorting then deleting the rows I don't want.

[Djtux]

6 minutes ago, zxcv64 said:

Thanks for the quick response. Yes, I have used the download function into EXCEL, but I thought if there was a function like the one I mentioned, then it would only need to be set once, and then in every future cycle any un-wanted stocks would then not appear on my list of stocks.

Just something to think about maybe at some point in the future, if possible.

Yes it makes sense, i was just mentioning a workaround.

For the filters, you mentioned a 'include' or 'exclude' list, a min price for stock. Do you see others filters as well ?

[StuC]

 

7 minutes ago, zxcv64 said:

Thanks for the quick response. Yes, I have used the download function into EXCEL, but I thought if there was a function like the one I mentioned, then it would only need to be set once, and then in every future cycle any un-wanted stocks would then not appear on my list of stocks.

Just something to think about maybe at some point in the future, if possible.

 

That would be nice to be able to save your filters on the web site.

[Djtux]

Just now, Stu Collins said:

This is what I do.  It's really simple and fast.  I can download into excel and have precisely what I want in 5 minutes by sorting then deleting the rows I don't want.

At the same time, if there is a way to make the workflow even faster, i'm open to suggestion depending on the complexity of the implementation.

[StuC]

2 minutes ago, Djtux said:

Yes it makes sense, i was just mentioning a workaround.

For the filters, you mentioned a 'include' or 'exclude' list, a min price for stock. Do you see others filters as well ?

NE IV and V would be nice as well I think.

[zxcv64]

1 minute ago, Djtux said:

Do you see others filters as well ?

This is just a 'nice to have' list off the top of my head - more for discussion and thought, as opposed to a requirement list.....ability to filter on :

- stock price (min, and max)

- whether stock has weeklies or not

- whether earnings date has been confirmed or not

I'm sure other folks will come up with suggestion too.

 

[mustafaoe]

Some kind of date picker (only earningsplay for selected week) could be also helpful. 

[Djtux]

Fyi, @Yowster reported by PM (thanks !) a bug in the calculation of the strangle credit (the blue dotted line) in some cases. It seems it's when the earning date is the monday before market open, which mean that the last trading day is the Friday of the week before. The strangle expiring that friday the week before is missing from the credit that we should receive. I will let you know when that bug is fixed. In the chart below, you can see that we should be able to sell the Feb 16 strangle to get a little more credit and so the dotted horizontal line should be lower. 

[Djtux]

1 hour ago, Djtux said:

Fyi, @Yowster reported by PM (thanks !) a bug in the calculation of the strangle credit (the blue dotted line) in some cases. It seems it's when the earning date is the monday before market open, which mean that the last trading day is the Friday of the week before. The strangle expiring that friday the week before is missing from the credit that we should receive. I will let you know when that bug is fixed. In the chart below, you can see that we should be able to sell the Feb 16 strangle to get a little more credit and so the dotted horizontal line should be lower. 

That should be fixed in production on the website.

If you see something something, you can report either here or by PM.

[Djtux]

On 1/29/2018 at 9:19 AM, Djtux said:

Yes, i'm aware of that for the source #2, sometimes it doesn't retrieve the nearest confirmed earnings because there are multiple confirmed earning date in the future. It's in my list of the 'to-do'

@RapperT That should be fixed now on the website.

[Yowster]

@DjtuxJust noticed you added the graph to the straddle table page, showing prior cycles implied moves and post earnings moves.   Great, thankjs for the quick turnaround on my suggestion!!!  No need to look at marketchameleon charts anymore.

[Djtux]

1 minute ago, Yowster said:

@DjtuxJust noticed you added the graph to the straddle table page, showing prior cycles implied moves and post earnings moves.   Great, thankjs for the quick turnaround on my suggestion!!!  No need to look at marketchameleon charts anymore.

Thanks, i've added that without communication. Those few days were a little bit busy. I have a bunch of other improvements suggestion that some members kindly provided, but i haven't got the time to implement everything.

For the Implied Move vs post earning moves, i'm not sure if the 'price effect' (return of stock close before the earning and stock close after earning) or the max move (return of the stock close before earning and the stock high/low after earning) is the best. Don't know if you have an opinion.

[Yowster]

1 minute ago, Djtux said:

Thanks, i've added that without communication. Those few days were a little bit busy. I have a bunch of other improvements suggestion that some members kindly provided, but i haven't got the time to implement everything.

For the Implied Move vs post earning moves, i'm not sure if the 'price effect' (return of stock close before the earning and stock close after earning) or the max move (return of the stock close before earning and the stock high/low after earning) is the best. Don't know if you have an opinion.

@DjtuxMaybe display all 3 for each cycle?

[SBatch]

2 hours ago, Djtux said:

Thanks, i've added that without communication. Those few days were a little bit busy. I have a bunch of other improvements suggestion that some members kindly provided, but i haven't got the time to implement everything.

For the Implied Move vs post earning moves, i'm not sure if the 'price effect' (return of stock close before the earning and stock close after earning) or the max move (return of the stock close before earning and the stock high/low after earning) is the best. Don't know if you have an opinion.

If only displaying one, I believe the max move is most relevant.

[Javier]

I really appreciate too, I was using other Web just to see your table in a graph. Agree if only displaying one "max move" is more relevant.

P.D. A simple mention is very appreciated to know something has happened. Thanks Yowster and Sbach for your heads up.

[Djtux]

I've added the max move in the barchart that already contained the straddle RV/Implied Move and Price Effect, i hope the chart is still readable.

[Yowster]

Looks good to me, thanks.

[Javier]

Nice work, can get rid the other web.

[Djtux]

21 minutes ago, Yowster said:

Couple of thoughts on GRMN...

• Average RV for prior cycles is so high because cycles prior to weeklies being available were included, so for about half the prior cycles the straddle did not expire the same week as earnings.

I always forget that, i will try to find a way to know when weekly options started for each ticker, and if i have that information maybe mark with a star the expiration date (in the RV charts or the RV table) and possibly the scanner.

It's not the first time i personally miss that point so i'm sure it must be annoying others too.

[Djtux]

I did a small change to the chart found on the "straddle table" page.

There is now an indication on when weekly options were introduced for that ticker.

Also it's interesting to note that for GRMN in July 2015 for example, the colors for the price effect (close-close) and max move are not the same, indicating the stocked jumped after earning before closing the day in the red.

 

 

Another example with KMX that we traded in the past :

 

Edited February 11, 2018 by Djtux

[Djtux]

I've added some changes :

• In the straddle RV charts : 1. there is a warning when weekly options were recently introduced. 2. There is the number of weeks between the earning date and the expiry date in the legends. 3. An option to force to use only monthly options for the long straddle legs
• In the calendar RV charts : 1. there is a warning when weekly options were recently introduced. 2. There is the number of weeks between the earning date and the expiry date in the legends. 3. There is the number of weeks between the short and long leg of the calendar. 4. An option to force to use only monthly options both for the short and long legs, in that case the nbweeks selection has no effect as i will select the first monthly after earning for the short leg, and the following monthly for the long leg.

[Djtux]

13 minutes ago, Djtux said:

For comparison purpose (still beta features and hidden), https://www.volatilityhq.com/backtester/return_matrix/?symbol=VALE&start_date=2016-02-12&strategy_type=40DCall&submit=Run+backtest

I don't select the same expiration (i select the expiration just after the earning date for the moment).

 

2 minutes ago, Sirion said:

Some minor feedback about this, generally wouldn't you do a color shift at breakeven (0) instead of 50%? Kinda threw me off how blue a lot of these were even though they're relatively good. 

Very impressive though, excellent feature for examining some of these CML trades and trying to see exactly how cherry picked some of the results are. 

You may have answered this in another thread, but do you only have EOD data access at the moment?

@Sirion For the heatmap color, that's a good idea. I will have to figure out how to do it with the charting library.

Yes only EOD data, it's going to be a while before i get intraday data because of costs.