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

Dissecting the Butterfly

Discussion in 'Options' started by Average Joe, Mar 19, 2016.

  1. Average Joe

    Average Joe Guest

    Some recurring themes that I've noticed is the idea of monitoring the legs of the butterfly to determine if an adjustment is required. How does one do that and what should one look for? I am a bit unsure as I don't intuitively see the usefulness of monitoring individual legs given that the greeks of the butterfly should reflect it. Or am I missing something?

    I'd love to hear what others have found useful and their approach in looking at individual legs.

    (I've heard Kevin Lee talked about this and most recently Chuck on the Trading Group March 8th video)
     
  2. vega4mike

    vega4mike Well-Known Member

    I think what Kevin is referring to is that sometimes when you look at the whole position you may not be able to easily identify the parts of your position that are no longer helping the trade & have become a theta drag. By looking at the individual legs of a fly (debit +credit spread combo), if the credit side can now be bought back for a very small price say 10c, then you may be better off buying back the credit spread & putting on another fly to get more theta into the trade or at the very least remove the risk since at this point risk reward of the credit part of the fly is no longer favourable. Looking at the positions greek, you may not be able to easily see this. At least this is how I see it.
    So its basically monitoring the debit & credit sides of the fly.
     
    Timo, Kevin Lee and Chuck like this.
  3. Average Joe

    Average Joe Guest

    Interesting. I haven't looked in to this but wouldn't looking at the butterfly theta also give you the answer? I mean the Greeks are additive so if the short spreads are not providing the amount of theta desired, the butterflies theta should reduce proportionately.
     
  4. vega4mike

    vega4mike Well-Known Member

    Agreed, however, does the reduction in theta tell you anything about your risk, experienced traders may see this & immediately realise that the credit side is now carrying far too much risk & make an adjustment, by getting rid of the risk and doing whatever to bring more theta to the trade, an inexperienced trader following rules may simply add more theta to the trade, leave the risk on because they are not aware of it , get more theta into the overall position & think everything is now ok.
    If your trading a simple fly & not a fly that has morphed into a complex position (M3, Rhino, etc), then looking at your greeks may quickly reveal the risk, but for complex positions, the risk is easily hidden.
    Also be aware that the greeks as they are lie to you (calendars are long vega, etc). After all the greeks we see apart from gamma are of the 1st order, there are second order greeks which may give you more information but are generally not available in the analytics we use.
    The closer you get to expiry the more unstable the greek get. I could go on & on about the unreliability of the greeks, but its suffice to say, you need a back up to your greeks when trading complex positions so that you can quickly see where your risk is & one way is to understand the individual legs of your positions, tedious! Yes,:(
    Of course its ok to simply rely on the greeks, but once is a while that hidden risk in that spread that you failed to spot in your complex position may just sprig a surprise on you one day...:eek:
     
    Rick, Kevin Lee and Chuck like this.
  5. SVL

    SVL Well-Known Member

    Yesterday I watched TG 2 episode from May 26, 2015 which convinced me even more that dissecting the butterfly is crucial for bringing your trading to the next level . I think spending considerable time studying how each individual leg performs over time under various market conditions is more beneficial when a simple back-trading .
    In the same episode Curtis Allen discussed 3 ways of reducing positive delta.
    Option A : rolling upper longs up
    Option B : rolling shorts down
    Option C: rolling lower longs up
    You could watch from 0:50 to 0.55 here:

    This is how Curtis’s M3 position looked on May 26, 2015
    2016-05-26 Curtis Allen M3  RUT Jun 15.png 2016-05-26 Curtis Allen M3  RUT Jun 15 Analyze.png
    I decided to study how each way of reducing delta performed over time .
    I tweaked the ratio of each spread in order to obtain similar amount of negative deltas. The risk profile of each adjustment looks as follows:
    2015-05-26 M3 adjustment A.png
    2015-05-26 M3 adjustment B.png
    2015-05-26 M3 adjustment C.png
    After doing this study I was actually very surprised that option A ( moving upper longs up) was a better option than other two. I believe it is primarily because Option A was the only one with positive Theta and negative Vega.
    2015-05-26 Curtis Allen M3 adjustment.png
    The above study was done for a range bound market with extremely low volatility. I plan to do similar dissecting studies for rising and falling markets as well as for different DTE periods and different volatility regimes. Hopefully after that I would be able to understand why a specific adjustment is better than other alternative options in a specific market situation.
     
    Last edited: Apr 11, 2016
    N N, Capt Hobbes and DavidF like this.
  6. Capt Hobbes

    Capt Hobbes Well-Known Member

    An important piece of the context is the purpose of the adjustment. When you cut negative deltas on the upside of an M3, an adjustment like that will stay on for a while. On the other hand, when you reduce positive deltas on the downside of a Road Trip faced with a risk of a down move, that is just a temporary fix for a day or two. After a few days, either the danger passes or you decide to pull the plug, at least on that particular trade structure. In the first case, positive theta is pretty much a must. In the second, even outright negative position theta won't hurt as much as the "slippery slope" on the left of a BWB could.

    Curtis' adjustment seems something more of the second kind. It's not very surprising that a positive theta negative vega position did better (or less badly) over a 16-day up move. But the important questions are: 1) Would he actually keep those verticals this long in this scenario? 2) How would these adjustments compare if the down move actually played out?
     
  7. SVL

    SVL Well-Known Member

    Dissect or not dissect ? In my previous post I only tried to demonstrate the importance of monitoring each individual leg.
    Yes, it is a very tedious work but it is the only way to understand why some adjustment is better then another.
    Curtis Allen said that he was most likely to make the adjustment C as he was concerned that if RUT goes up, adjustment A would hurt him the most.
    Without dissecting the butterfly we would not be able see that it was not the case.
    Only after doing hundreds of similar exercises and spending countless hours, we would be able to reach the proficiency in trading complex option positions. This is the ongoing theme in many interviews of John Locke with the successful traders of the month and that particular TG 2 session on May 26, 2015.
     
  8. SVL

    SVL Well-Known Member

    Another excellent advice was given by Ron Bertino during TG 2 session on Apr 12, 2016 ( 0:15 to 0:22) about monitoring a broken wing butterfly as 2 individual components : a regular BF and a credit spread. Once 80 % of credit spread earned, it is time to make an adjustment.
     
    Last edited: Apr 16, 2016
  9. Gabor Maly

    Gabor Maly Well-Known Member

    Very interesting analysis and thanks for posting this SVL. It does make sense to test this in multiple environments, while Option A did come out well in this relatively muted "rebound", you may get different results in a scenario that Curtis in the video was afraid of which is a quick face ripping move up after a sustained move down...in that case deep ITM longs will get crushed and probably suffer more than option B or C (and the profile of A you have posted seems to be in line with that). Again to be tested, I do however at the moment try to keep my longs closer to the money once we get a sustained move down.
     
  10. Al G.

    Al G. Well-Known Member

    In trading the Road trip Dan/Tom reccomend "layering' the trade every two weeks. The idea is that the RTT has positive expectancy and the more of them you put on the better. I really like the concept of layering (also introdduced with space trip trade) , but in low vol environment in two weeks the original RT might not have moved much so in essence you have the short of both RT strikes relativel close to eachother.

    Can anyone share how they go about layering, can we use price moves as a guide of when to layer eg if price moves 2 ATR than put another RTT one. I would like to know what rules of thumbs other traders utilizee in their layering technique.

    Thanks
     
  11. Capt Hobbes

    Capt Hobbes Well-Known Member

    I'm not questioning the importance of detailed adjustment analysis. What I'm saying is an adjustment should be first of all viewed in the context of the problem it's intended to fix. That dictates which of its properties are more important that others. Further, it's good to begin the analysis of an adjustment by reviewing a rough theoretical model of the greeks. Individual leg monitoring is fine for revealing where real life doesn't follow the model, but without a model we might see just the trees and not the forest.

    Here is a way to look at Curtis' example.

    His first priority is to fix the delta which is too positive so a continued move down could hurt. This is a stopgap solution until it becomes clear what's really happening. He is considering rolling up the upper or the lower longs, or rolling down the shorts. All of those are different names for the same thing, adding a bearish vertical. The only difference is its position relative to the money. We want to compare what these three different verticals add to the existing structure (so strictly speaking, we are not dissecting anything).

    Here are greek profiles of a bearish vertical with 33 DTE, not quite the same as the example, but good enough for a ballpark understanding. The spread is 10 points wide, but the general shape for a 20-point one is not very different. It's helpful instead of showing three curves for three locations of the vertical to show just one curve with the position of the market relative to the spread marked with letters A through C. I erased price labels to make it less confusing. Gridlines mark roughly 1% moves, and the two vertical dotted lines show where the strikes are. With adjustment A the market is about 4% below the lower strike, with B it's about 1% above the upper strike, etc.

    vs-delta-ann.png

    Immediately we see that adjustment A is very different from B and C in terms of delta, because the market is on the opposite sides of the dip in the curve. Which means even if the initial adjustment delta is the same, as the market moves down adjustment A will pump less negative delta into the position, while B and C will pump more, and vice versa. Clearly, for our purposes it's better if the added negative delta increases as the market moves down and decreases as it moves up, which makes B and C superior to A. (All of which in a long way of saying that the gamma of A is negative and of B and C is positive, and adding some positive gamma to the original position's negative gamma is good). Comparing B and C, we can also say that C maintains this desirable property of positive gamma (downward slope going left) over a wider range on the downside, and B does so on the upside (upward slope going right). So overall, from just the delta viewpoint both B and C should give us a flatter t+0 line than A, and C will do better for a continued move down, while B will be slightly better for a reversal up.

    vs-theta-ann.png

    Theta is not important for our stopgap adjustment. However, the theta graph certainly explains the difference in behavior over a longer holding period. For a longer period, it's important to keep in mind that these curves morph with time and IV. The curve above would suggest that for adjustment C a 2% (2 gridlines) move up would change the theta from negative to more negative. However, in SVL's study that move plays out from 26 to 7 DTE. Over that time the curve contracts a lot along the X axis, so we should expect theta to go more negative and then less negative, which is what the study shows.

    vs-vega-ann.png

    The vega graph again shows that A is pretty much the opposite of B and C. Again, if more downside is our worry, then an increase in volatility probably goes along with it. For that scenario, the greater positive vega added by C is a plus.

    Again, these are just general illustrations, far from any sort of precision. But hopefully they make it pretty clear what in general to expect from each of the alternatives, before even collecting any data.
     
    Last edited: Apr 17, 2016
    N N, SVL, meyer99 and 2 others like this.

Share This Page