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

M3 vs BWB - A Reductionist Approach

Discussion in 'Round Table Presentations' started by Kevin Lee, Jun 19, 2016.

  1. Kevin Lee

    Kevin Lee Well-Known Member

    Hi all,
    Many people have questions about M3 vs BWB. I created a video to show the differences between the two trades in a unique way. I call it a reductionist approach. I find it useful in helping me understand the two trades better. I hope it'll do the same for you.

    In addition, the video will also show an example how to use 2nd order greek, the vomma, to analyze and compare the two positions.

     
    Last edited: Jun 19, 2016
  2. DavidF

    DavidF Well-Known Member

    Excellent Kevin, very informative, thank you.

    Something I´m not clear about, you´re changing volty % to get a more accurate prediction in OV as you mentioned that IV is static in OV´s model. I was always under the impression that OV built in an estimated change in IV that would correspond to changes in price in their model. However I did start to wonder last week when the model predicted I would be $8k up with an SPX move to 2070, but I was in fact about $4k down. When I plugged in a 8 point increase in volty% it was much more accurate.

    If this is correct we should always manually add or subtract some points in the volty% box to get a more accurate prediction for P/L? Maybe the issue last week was that VIX spiked much more than the move in SPX in relation to past moves.
     
    Last edited: Jun 19, 2016
  3. Kevin Lee

    Kevin Lee Well-Known Member

    David,

    This is an excellent question. Remember the 3 different type of shifts that the IV curve goes through? OV does model the horizontal and vertical shift. But it doesn't model the skew change. However, even for the vertical shift, the amount of IV drop or increase may be difficult for OV to get it right. They can only use an average amount based on historical data. Therefore, it's prudent to compensate some more when we look up or down the T+0 line.

    Remember though when we add or subtract volty%, OV compensate each strike the same amount. In reality, there will be a skew change. Unfortunately, that effect cannot be simulated by the software. For that, we need to be familiar with how skew change affects the position and then manually compensate for it. In addition, when time passes, the skew will naturally steepen some more. So we have to keep all these nuances in mind.

    For last week, it's a difficult one. IV went up a lot more than normal for a relatively small movement in the underlying. Therefore, OV definitely under forecast. In addition, the skew steepened. As a result, all our T+0 lines behaved unexpectedly. That's hard for any s/w to get it right.
     
    Chuck and DavidF like this.
  4. DGH

    DGH Administrator

    Excellent, Kevin. Now, if only TOS would give us the choice to incorporate stochastic volatility into their pricing model. I have tried modeling strategies using GARCH projections of future volatility (one method of calculating stochastic volatility, as you know) and smile/skew modeling using varying degrees of kurtosis of the smile curve. This approach can be helpful in attempting to choose the "right" volatility to use in calculating standard deviations and probability cones. Interesting stuff, but I did not find these projections useful on a routine basis, at least for retail traders like myself, but can be useful in trying to predict "what if" situations. The Big Boy firms have their own quants, of course.
     
  5. Kevin Lee

    Kevin Lee Well-Known Member

    Agree. I don't think we need to get it numerically correct though. That's not useful for our day to day trading. However, what I find useful is how the volatility changes might affect the T+0 line. Then I'd mentally compensate for the greeks and/or the shape of the T+0 line. That compensation is just a very rough estimate. Nothing numerically accurate. I use that to guide my adjustment decisions.
     
  6. DavidF

    DavidF Well-Known Member

    That makes sense Kevin, thanks again! Obviously lot of moving parts.

    You mentioned in the earlier presentation the poor performance of long OTM calls with a crush in volatility, which is of course correct. However I notice that it depends a lot on the DTEs and if the degree of the move up.

    For example say you enter an M3 on the RUT Feb 11th where it´s at 953 after the major down move an IV spike. You get 190 deltas with either 2 x front week 870 calls that you roll every week or you choose 4 x April 960 (ATM calls). The latter has positive vega of around 600, the former almost zero. So you´d expect the ATMs to get IV crushed. However the move up is so quick and the calls are far enough out that the gamma on the ATMs helps them outperform on the whole move up. By March 3rd the 4 calls are now giving you 370 deltas and are worth $31k whereas the DITM calls can obviously only ever provide 200 deltas and are only up $23k.

    When it stops moving it could obviously be problematic if it´s outside the BF but these moves up after a IV spike can be so violent that they´re worth considering, esp. as downside risk is also less (cheaper).
     
  7. Kevin Lee

    Kevin Lee Well-Known Member

    You are right David. If the underlying explodes upward, firstly the skew may not steepen as much as expected. Secondly, the delta gain will overwhelm the IV crush even if it happens.

    However, it's hard to predict when market will spike upwards. It seldom does. So in the long run, my experience is that having a OTM call hedge is not a good strategy. Unless the call hedge is meant only to tie over a very short period of time, ie couple days.
     
  8. DavidF

    DavidF Well-Known Member

    Hi Kevin, I´ve been thinking about your M3 vs. BWB reductionist approach and would appreciate your thoughts on one aspect.

    Say you wanted to compare an M3 vs. a BWB for this October expiry.

    Let´s say the M3 a 5-lot 2025/2125/2225 with a 2025 call.
    The BWB is a 2025/2125/2200. The delta/theta are relatively close and you can reduce the difference to a 5 x 2225/2200 credit put spread vs a 2025 call.

    However if you compare the overall positions you will see the starting points are very different and the M3 has more absolute downside risk and the is closer to the middle of the tent vs a roll-down the long puts BWB.

    Wouldn´t a fairer comparison be to roll the short strikes up in the BWB to get 2025/2150/2225 BWB? Of course the short strikes are not the same but the absolute risk is similar, the graphs are comparable and the reductionist approach is then a 2025 call versus a short 10 x 2150/2175 put spread.

    You´re now comparing a 100 strike balanced, versus a 125/75 if you roll the shorts up and a 100/75 with your original BWB set-up.

    As you can see from the graphs below of the reductionist comparions the pros and cons of an M3 vs. a BWB depend on whether you roll the short stikes up vs. taking longs down. From the graphs of the overall positions I would argue that rolling the shorts up is a fairer comparsion with an M3 with respect to positoning within the tent and max risk.

    With respect to comparison of a M3 vs a BWB when there´s a steep or shallow put skew, not sure if you´d be aggravating the effects of the BWB by selling closer to ATM) versus having the ITMs closer to the money. But it´s clear in a market like the current one (comatose) you can get a lot more theta by having your shorts closer to the money vs. long calls.

    You can see the comparions here, straight call, roll down the long puts BWB and a roll up the shorts BWB with obvous pros and cons. Also interesting to note the highest positive vega (in theory) below the money is with the rolled-up puts.

    Screen Shot 2016-08-27 at 11.46.42.png Screen Shot 2016-08-27 at 11.46.45.png Screen Shot 2016-08-27 at 11.46.47.png
     
    Last edited: Aug 27, 2016
  9. Kevin Lee

    Kevin Lee Well-Known Member

    Hi David,

    The reductionist approach is a short cut way for me to analyze something quickly. It works in some situations but it might make it even more complicated in others. The example I gave during the presentation is the configuration that I commonly encounter, ie morphing back and forth between M3 and BWB depending on the IV skew. Therefore, it's only shifting the right leg to form the BWB and leaving the rest of the legs unchanged.

    Here's how I would compare the example you showed. The T+0 lines are pretty close. This only requires the shifting in of the right leg, which is a bullish vertical adjustment.

    upload_2016-8-27_7-5-44.png

    upload_2016-8-27_7-5-58.png
     
  10. DavidF

    DavidF Well-Known Member

    Ok thanks Kevin. My post was more in relation to my realisation (using your reductionist approach) that a comparison of an M3 and a BWB depends on what factors you keep constant (e.g., short strikes or max. risk or postion in tent at outset). Anyway, useful way of analysing set-ups and adjustments.
     
  11. Randy Schwartzenburg

    Randy Schwartzenburg Well-Known Member

    One aspect of the diff which doesn't have anything to do with the greeks is execution. One reason I like the BWB is because it can be taken off with a GTC order at a set profit verses the M3 and having to deal with exiting a long call when the market moves unexpectedly after the butters are taken off. This applies to placing each trade too.
     
    Last edited: Aug 30, 2016
  12. DavidF

    DavidF Well-Known Member

    Agree Randy, in fact I think the only bad expereinces I´ve had with an M3 have been execution. Worst (amateur) mistake I made was not peeling off calls/BFs on an SPX M3 position, took 5 calls off and quickly tried to close the flys but ECB made an annoucement that sent SPX flying up, saw close to 10K profit (that I really had to work for through Brexit) reduced to about 1K. That´s of course an amateur mistake, taking off calls first and not peeling off. But even then I´ve often found I eventually get filled on the flys just as the market turns and am left chasing the falling market on the calls.
     
    Trader G likes this.
  13. N N

    N N Well-Known Member

    Speaking of execution, I noticed IB has a 'market price' for marking the flys and most on here are quoting the mids. As has been shown before the mid price can jump all over the place intraday and was curious how people are pricing the flys where they ACTUALLY might trade OR how IB is calculating the 'market' price?
     
  14. Randy Schwartzenburg

    Randy Schwartzenburg Well-Known Member

    I just put my profit target in and drink cold beer on the back porch until it fills. :)
     
  15. Steve S

    Steve S Well-Known Member

    There's no substitute for getting set up to track the fly price recent history as has been discussed (and tools provided and/or explained in detail) many times in this community ... but for IB quick-and-dirty just pull the real-time leg data, including sizes, into your spreadsheet and add the fly formula ... when the leg prices are off it's usually visible in the spread and/or sizes for the leg(s) in question and you can adjust by nudging the leg bid/ask(s). Pulling IB data is easy because the ActiveX spreadsheet sample they provide is VERY full-featured (has sample code for absolutely everything) so you can easily modify it without investing much time or energy.
     
  16. Kevin Lee

    Kevin Lee Well-Known Member

    I fully agree with Steve. I learned a lot by just capturing and monitoring how butterfly prices behave in up market, down market, steep IV, flat IV, market in fear, etc... It's really impossible to put everything down in guidelines to follow as some traders have suggested that I do.

    However, I would suggest tracking butterfly prices, IV skews, etc.. only to intermediate and advance traders. For beginners, I think the priority should be on learning to trade the plan. Otherwise, you'll be juggling too many balls before you're ready.
     
    Paul Demers likes this.
  17. Steve S

    Steve S Well-Known Member

    Agree 100% with Kevin's points ... if you're not a coder or an Excel power-user then these data things tend to be a big hassle and it's much better to focus your best energies on trading the plan ... I'm only posting again because it just struck me as funny (ha ha funny) that for a generic trader contemplating a generic trading instrument it's usually regarded as natural and necessary to first "pull up a chart" ... so a trader new to options spreads would naturally first say "how do I pull up a chart on this?" before even considering a trade ... but, surprise! This is a big fat hassle for complex options spreads and most traders give up and "trade blind" - kind of insane when you think about it.

    With Interactive Brokers you can pull these charts through the API by requesting bid/ask historical data for each leg and processing/displaying yourself in code ... maybe you can do it in TWS without the API, but I don't think so as bid/ask/mid seem to be unavailable for spread charts. I'm surprised you can't do it in TOS, or can you?
     
  18. Kevin Lee

    Kevin Lee Well-Known Member

    Steve,

    In Tos, we can only get live data through RTD. No historical data. Question - for IB, are you able to get historical end of day or intra day data through API ?
     
  19. Steve S

    Steve S Well-Known Member

    Kevin, the IB TWS API historical data capabilities are excellent so long as you don't request too many data points at one time (but the limits are not onerous for normal trading purposes), and so long as you don't make too many requests in a short time period ("pacing violation", request denied). I'm not sure about tick data (I thought you could get it, but it's not in the current documentation), but you can request bars from 1 second up to monthly, and you can request bid/ask/trades/mids. You get the historical data right up to the last tick, no delays in serving the data. My main caveat is that I've never used the API for historical option data, just stocks and futures, so I can't say what kind of problems might arise. I prefer doing something along the lines of what you do - dump the current chain data to text files at intervals and use my app to create the charts I want in Excel.
     
  20. Kevin Lee

    Kevin Lee Well-Known Member

    Do you mean IB allows you to dump chain data for a specific time of a particulate date into text or csv file like OV ? That's fantastic !
     

Share This Page