Quantcast
  1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Weighted Vega

Discussion in 'Options' started by tom, Jul 24, 2015.

  1. tom

    tom Administrator Staff Member

    Ron Bertino's presentation on Weighted Vega was a "must see." If you haven't seen it yet, go here:



    Ron got a call from Karen at OptionVue who said she loved the presentation and was forwarding it to Len Yates to consider adding weighted vega to OptionVue. That would be excellent if Len would do that. Bill Ghauri already added weighted vega to his TOS tools that he provides to his Alpha Alert subscribers.
     
    Last edited: May 16, 2016
    GreenZone likes this.
  2. Kevin Lee

    Kevin Lee Well-Known Member

    It is indeed a great presentation. Ron - thank you very much for the great work. I had first hand experience with what Ron presented. What he described is absolutely spot on.

    Calendar was the first strategy I learned. After traded for a while, I realized calendars just weren't behaving the way that it is supposed to. I watched a class taught by Mark Sebastian and I learned the concept of weighted vega. Suddenly, it became clear to me the odds were so stacked against me. What I've learned about calendars weren't entirely correct. I did continue to trade calendar for a while after that but I switched from trading it for monthly income to trading only when the horizontal skew presents good opportunities.

    For those interested, here's a pretty old video by Mark on weighted vega and how to deal with calendar. http://video.theoptionclub.com/mark-sebastian-on-calendar-spreads.html
     
    Last edited: Jul 24, 2015
    GreenZone, RayM and M L like this.
  3. DavidF

    DavidF Well-Known Member

    Yeh fantastic presentation. I actually started watching it live but stopped when I thought "this is a calendar issue". Watched it all the way through now and see that it´s definitely not only relevant for calendars. Coincidentally had noticed that my Sep M3 trade was positive despite a 2-3% down move in contrast to what my vega number would have predicted. Thanks for the education.
     
    GreenZone likes this.
  4. vega4mike

    vega4mike Well-Known Member

    Excellent Presentation & Explanation on how vega impacts option positions.
     
    GreenZone likes this.
  5. GreenZone

    GreenZone Well-Known Member

    I'm happy to hear that you guys liked the presentation.
    Hopefully this will get the concept of weighted vega "out there", and people will be able to take their trading to a new level.
    I'm looking forward to seeing some creative traders put new trades together which can take advantage of these properties, similar to the way that Dan Harvey took the research done by Brian Johnson and started creating new trade types (eg Weirdor variant).

    I also wanted to say a big "thank you" to Tom and Jim, for running a fantastic trading community which is truly open to discussing new ideas.....even if the ideas may be a bit controversial.
     
    Randy Schwartzenburg, Bruno and RayM like this.
  6. kosta

    kosta Guest

    Hey Ron,

    Firstly, the Roundtable discussion / presentation was phenomenal. You have saved a great many future traders a great deal of pain (myself included:). My background was fixed income analytics (complex and difficult to price instruments), before going out on my own trading spot OTC. Survived that trial by fire and am now diving head long into options; in particular reverse calendars.

    I have a question, that I hope you can help with.

    Everyone I speak to tells me it's silly-hard to extracts from that short call or put once the long expires but your presentation has now made me consider otherwise.

    If both legs are ATM European style, and the DTE's are say 700 and 1000 for the long and call respectively (e.g DAX or STOXX 50), shouldn't I be little troubled by IV for the simple fact that weighted vega will likely be neutral and I actually want IV to climb so my gamma and delta become such that my options start behaving like futures ? If they move far enough, I would simply close out and keep the bulk of my credit right? Or am I missing something? (E. Term structure of IV for example)

    Kosta
     
  7. GreenZone

    GreenZone Well-Known Member

    Hi Kosta

    Thanks for the feedback. I'm glad to hear that you enjoyed the presentation.

    Ok, so let's get down to business...
    You say that you're getting into reverse calendars, so let's start there.
    For the sake of others who may not be familiar with the trade, let me give some background.

    A reverse calendar is a long gamma trade (the T+0 has a U shape). In that sense, it's similar to a long straddle.
    Here's a 5 contract reverse calendar in blue (longs are at 31 DTE and shorts are at 49 DTE), and a 1 contract long straddle in green (with longs at 49 DTE).
    You can see that the T+0 is fairly similar on both trades, and they are long gamma.
    Both of these trades are going to make money if you get a big price move of the underlying in either direction.
    [​IMG]
    Now let's compare the greeks:
    [​IMG]
    Some things to notice:
    • The margin on the reverse calendar in a RegT account is insane (almost $221K) vs a cost of just $7K for the long straddle; this is due to the fact that you have a short option in a back month, so the broker is assuming that your front long option will expire and you'll be left with a naked short. This implies that reverse calendars are realistically only viable within a Portfolio Margin (PM) account.
    • the deltas both start off pretty flat, so no difference there
    • the long gamma is almost identical between the two trades
    • both of these trades have negative theta, which means that time is working against you; and notice that the theta on the long straddle is 3 times larger than the reverse calendar
    • the vega on the reverse calendar is negative, whereas on the long straddle it's positive
    So we already have a few interesting things to note here:
    • If we have a PM account, then we can consider doing reverse calendars, otherwise they are not viable in a RegT account
    • With a reverse calendar we get the same amount of long gamma (this will make us money if price moves significantly) at one third of the theta decay of a long straddle. Think of the long gamma being the "product" you are buying, and the theta decay as the "cost" of that purchase. You are therefore buying the same amount of the "product" (long gamma) at 1/3 of the cost (theta decay) if you use the reverse calendar compared to the long straddle.
    • Forgetting about weighted vega for a moment and just looking at the raw vega, the reverse calendar is negative vega which means that an upside move (which typically results in lower IV) will help the position more than a downside move (which will typically increase IV). In contrast, the long straddle is positive vega, meaning that the position favors a downside move more than an upside move. Therefore, one of the things to think about when trying to decide between a reverse calendar and a long straddle is if you have a directional bias. If you think it's more likely that price will break out to the upside, then a reverse calendar may make more sense. Of course, you need to take the other factors into consideration as well (eg theta decay).
    (I'll continue this in a post below this; the server is running into errors)
     
    Last edited: Aug 1, 2015
  8. GreenZone

    GreenZone Well-Known Member

    Next we'll look into the days to expiration to choose.
    If you recall from my weighted vega presentation, I explained that as you move further out in time, your gamma will decrease.
    In your post you mentioned using 700 DTE for the longs and 1000 DTE for the shorts. Now that's WWWAAAAYYYY out in time.
    Without even looking at the risk graph, you should realize that when you go that far out, the gamma is going to be a lot smaller.
    Having small gamma is typically considered a good thing, but this is only if you are trading a conventional negative gamma (and positive theta) trade.
    But in this case, these trades are trying to make money from big price movements of the underlying, and therefore take advantage of positive gamma.
    So when we do these types of trades we actually want our long gamma to be as big as possible, for the smallest cost (theta decay).
    This is why most people who try to make money from these types of trades will normally choose a DTE of around 60 days. This tends to be a good balance between giving the position sufficient time to make the move, and maximizing the amount of long gamma you are buying.
    Ok, so now let's look at exact numbers, for the example you mentioned.
    I used a DTE of 686 for the longs and 868 for the shorts. That's as far out as my software can go.
    http://screencast.com/t/pAN2aVRGx4m
    Take a look at the gamma. It's 18 times smaller than the first example.
    This is not good when you're trying to make money from big price movements. We want the biggest amount of long gamma for the smallest theta decay cost.
    One of the things you can do is to look at the ratio between gamma and theta. You want gamma to be as big as possible and theta to be as small as possible, so you would look at the ratio of gamm/theta. The bigger the end ratio, the better the "deal".
    In the first example you would have: 0.89 / 23.61 = 3.77%
    In the second example, you would have: 0.05 / 7.4 = 0.67%
    So you can see that the first example is giving you a better "deal" (you are buying the most amount of long gamma for the cheapest theta decay cost).

    You will likely also have liquidity issues by going out so far in time.

    Last but not least, we have weighted vega.
    If you were to use Nassim's formula, you are correct in suggesting that the true vega would be much closer to neutral than the numbers shown in the raw vega. It would still be negative, but it would be smaller in size than what any software suggests.

    But hopefully from all of the details shown above, you'll realize that there are bigger issues here than the weighted vega concerns.

    There's nothing wrong with reverse calendars, but I would suggest that you backtest using much smaller DTEs.
    Try a 30 day / 60 day, or a 60/90.
    I wouldn't go out further in time than that.

    I hope that's helped clarify things a bit more for you.
     
    Walker and RayM like this.
  9. ACS

    ACS Well-Known Member

    Would a reverse calendar be more margin friendly with futures options that use SPAN rather than RegT?
     
  10. GreenZone

    GreenZone Well-Known Member

    Yes, you're correct.
    Doing reverse calendars with futures options, which have SPAN margin, should also be viable......even if you have a RegT account.
     
  11. GreenZone

    GreenZone Well-Known Member

    Hey guys/gals, I've got some people who are trying to vote down my weighted vega presentation on YouTube.


    If you've watched it, and enjoyed the presentation, please give it a thumbs up.
    And if you want to leave a feedback comment, then I wouldn't hold that against you ;)

    You'll need to watch it on YouTube, rather than embedded in this forum post.
    To do that, click on the play button first, and then click on the YouTube icon on the bottom right.
     
    Last edited: Aug 4, 2015
    tom likes this.
  12. Gabor Maly

    Gabor Maly Well-Known Member

    Done.
    By the way what was their argument against your weighted vega concept? Just curious...I don't want to open up a they against us discussion just wonder if they had any valid, solid arguments. You have done an excellent job demonstrating and explaining this subject and I am sure we all say that without any bias towards any community.
     
  13. Paul Demers

    Paul Demers Well-Known Member

    no solid argument against it as far as I could tell
     
  14. GreenZone

    GreenZone Well-Known Member

    There were a grand total of zero arguments shown.
    Just generic comments such as "anything can be proven with backtesting", "I was in the pit for 25 years and I've never heard of weighted vega", "that guy isn't an experienced trader", "who here has ever used weighted vega before?", etc, etc.
    They claim that in the next 2 to 3 weeks, they'll create a presentation of their own, insinuating that they will "disprove" the validity of weighted vega.
    I look forward to seeing it.......although the most likely outcome is that nothing further will be done on their part, which is a shame.
     
  15. tom

    tom Administrator Staff Member

    When Dan says he was in the pits for 25-years, he's not doing the math correctly. He started as a runner in 1982 and left the pits in 2004 to start mentoring with OptionVue. My math says 22-years.

    Dynamic Hedging by Taleb was published in 1997 so surely someone in the CBOE pits must have noticed it before Dan left in 2004.
     
    GreenZone likes this.
  16. GreenZone

    GreenZone Well-Known Member

    I did a bit of a follow-up session on weighted vega in today's (Aug 4, 2015) Trading Group 2 meeting.
    I also explained the difference between a reverse/short calendar and a long straddle, and lastly I covered some aspects of gamma scalping.
     
  17. Paul Demers

    Paul Demers Well-Known Member

    During the Trading Group 2 meeting I posted in the chat that in fact Optionvue does reflect some form of weighted vega and I thought I would show an example of how I came to think that way. This first image is an ATM put calendar. I have set OptionVue to mimic what you would see in thinkorswim. I set the Delta/Gamma to Standard model and set the Selected Volty Model to EIOIO. I never have the Use combined call/put checkbox checked and that is reflected in all the images.

    Cal1.jpg

    In the next image I increased the vols by 5% to make the presentation a little more dramatic.

    Cal2.jpg

    The last image shows changing the OptionVue settings to reflect True delta/gamma and using their Variable model.

    Cal3.jpg
    I am not sure how accurate their calculations are but you can see the dramatic difference in the T-0 line and I believe it is close enough to get a better feel for what the expectations will be with a vol spike.

    Hope this helps

    Paul
     
    GreenZone likes this.
  18. GreenZone

    GreenZone Well-Known Member

    Hi Paul,

    I tried to address your question during the session, but perhaps my explanation didn't come through clearly. It's a tricky subject.

    After my weighted vega presentation, I got in touch with Brian Johnson to see if he had feedback over the presentation and/or information about any existing work he might already be engaged in with regards to implementing weighted vega into OptionVue (since I had heard rumors that he was already engaged with OV over wanting to implement weighted vega).

    Here is a quote from one of the emails he sent me on July 23, 2015 (he gave me permission to post it):
    I've also been contacted by OptionVue staff asking that I send them all of the materials I presented so they can review it, and consider potentially making changes to the software.

    So you are correct in saying that the T+0 models are quite different between EIOIO models (such as TOS) and OptionVue, but this is a function of CEV calculations across the term structure (ie CEV is calculated across each leg of the calendar), rather than being a function of weighted vega already being implemented in the OptionVue software.

    Here's another example of what I mean.
    If I setup a calendar in RUT with the short at 16 DTE and the long at 163 DTE, using 10 contracts, then the position greeks show a positive vega of 2236.
    calendar.jpg
    The software is therefore telling us that if IV goes down by 1 point, then we will have a loss of -$2.2K
    We know this is not the case, due to weighted vega.
    This is what I mean by the fact that weighted vega isn't reflected in the T+0 nor in the position greeks.....yet.
     
    Kevin Lee likes this.
  19. Kevin Lee

    Kevin Lee Well-Known Member

    Whenever I hear people say I was in xxxx for yyyy years and things were never done this or that way, I'd recommend that they watch this ....

    I'm a firm believer that trading is fast becoming a data science instead of an archaic craft of yesteryears, it doesn't matter if you have a century worth of experience, it's always facts and evidence that matter. Ron - keep up the good work !
     
    Last edited: Aug 6, 2015
    Hugh, Gail, Gabor Maly and 3 others like this.
  20. Kevin Lee

    Kevin Lee Well-Known Member

    Ro, I agree with your statement. Analyzing T+0 by increasing or decreasing the volatility steps in either TOS nor OV doesn't work because the increments are applied equally to all the legs instead of on a weighted manner. Actually that applies both to the horizontal as well as vertical spreads. Before i learned about weighted vega I used to do the increase and decrease method during my fire drill as taught in SOM. I realized that doesn't work, neither for calendar nor butterflies. In my opinion it is not even a close approximation to reality because at times the results are completely the opposite, ie increasing or decreasing the vol steps to analyze down or up market movements actually made the analysis worse.

    I truly hope there is a way to implement this in OV but i am speculating that it will be difficult because the weightage is a function of the IV skew and that is dependent on market forecast of the IV change. I'm not sure if there exists any theoretical model that can do this well enough, but maybe there is. Alternatively, OV might have to do a statistical analysis of past data to come out with an empirical model instead, but that would have to be done on a per underlying basis because i think the skew shifts differ from one symbol to another.
     

Share This Page