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

M3 with Weekly Calls - Discussion of DITM Call Time Premium

Discussion in 'Options' started by AKJ, May 19, 2016.

  1. AKJ

    AKJ Well-Known Member

    There was a very healthy discussion during Tuesday's Trading Group 2 recording that I was unable to attend (I sadly am seldom able to attend) about the use of weekly calls in an M3 trade. I wanted to add to the discussion with a tidbit that may clear up the question of whether or not these weekly calls can and ever are executed with "negative" time premium.

    It is true that you can have these trades executed at "negative" time premium if you are looking at the T.Prem column in OptionVue; this is because OptionVue defines the time premium for calls as [Market Price] + [Strike] - [Spot]. By following this formula, you will come up with a negative number occasionally for DITM calls with relatively few DTE's.

    In reality though, the time premium of a cash settled European option is [Market Price] + [Strike] - [Forward]. When the dividend yield of the index is higher than the risk free rate, as it is now, the Forward price will be below Spot. Calculating it this way, I highly highly doubt that you will ever be able to execute a DITM call trade with true negative time premium. This is a pure arb play if there is a listed future and a high probability risk-arb play if there is no future, and computers are likely to gobble it up.

    So when Tim states that he will sometimes execute these trades with "negative" time premium, I do think based on what OptionVue is telling us, that is correct. But this is no free lunch. The actual time premium is still positive.
     
  2. Boomer34

    Boomer34 Well-Known Member

    I am investigating the weekly M3 call...
    For those that use this approach, how far out do you buy it, and when do you roll it?
    Are you rolling it to the next week each time? (Ex: You enter at 56DTE and exit at 28DTE...would you have had 4 different weekly calls during that 4 week trade?)

    Thanks for any clarity and direction!
     
  3. Tim

    Tim Well-Known Member

    Andrew, I understand the theory and the calculations. However, I have been filled at prices on DITM calls that are below intrinsic value (for a buy). I did it again yesterday.Regardless of what OV says here is a made-up example (because I don't have actual market prices from yesterday when I filled this order but this indicates what happened).

    RUT: 1101.35
    Buying 2 lot MAY4 1000 call.
    Now, intrinsic value is 1101.35 - 1000 = 101.35. There should be some time premium in the option as well, so it "should" be quoting for a little more than 101.35. I was able to get a fill for less than 101.35, by about 15 cents if I remember right. Typically I can get fills for 101.10 to 101.35 in these conditions. The deeper ITM it is, and the fewer DTE it is, the easier it seems to be to get these fills. Fills like this usually happen within seconds of placing the order. Does this happen because the market shifts by .15 in the short time the order is in? Possibly, but it seems like this isn't the only explanation.

    OV does show misleading information in that a) the time value shows as negative sometimes by 1.00! I think however that when it shows negative it is an indication that there likely isn't any and you can go to the market and see what bid, ask and quotes are. And b) OV shows negative theta when you add this option to your position. I believe it is using theoretical calculations. If you don't buy any time premium, there isn't any theta.

    as a side note, if you are using TOS quotes to feed OV, set your model to Bid+Ask/2 to get mids that match what TOS shows. If you are using their quotevue, I belive the model is (10Bid+10Aked+Last)/21 to show the same mid values.
     
  4. Tim

    Tim Well-Known Member

    Boomer, there really should be a webex whole discussion about how we can use weekly calls. It took a couple weeks of doing it and talking with others for me to figure it out. But,

    I generally buy it one to two weeks out when I open the position. Selecting the call for little to no time value. It generally will be deeper ITM than a 90 delta monthly call, and generally but not always costs less. If usually has 99 delta or so as reported in OV.

    Then, in the week it expires I will roll it out at least one week and sometimes two. You have to look at the prices. The idea is, to make this roll (which is selling near week and buying further week which is a calendar spread), for even or even a small credit. If you do this, you continue to not buy time premium. Much more to talk about, but DO NOT FORGET to do the roll and let it expire. I did once and went through a weekend with no calls on my M3. That was really uncomfortable and is the only time I ever bought a future on the weekend as a hedge.
     
  5. AKJ

    AKJ Well-Known Member

    Hi Tim,

    Thanks for sharing your example from yesterday. I understand what you are saying, but want to revisit my point about using the Spot versus the Forward in determining time value, theta, etc. of cash settled European options. A deeper dive into the difference between spot and forward pricing will hopefully clear up the concept.

    In your example, you say RUT is 1101.35, so therefore the intrinsic value of the 1000 call is 101.35. This is wrong. Spot RUT is 1101.35, but the forward price of RUT is LOWER than 1101.35. At this moment, the June RUT future is trading ~1.5 points below RUT, so for argument's sake, let's say the June RUT forward price is 1100. For a June 1000 Call, the intrinsic value is 100, not 101.35. In this case, if you were able to execute a DITM call at 101, the true time value of this is positive $1, not negative $0.35.

    Now the MayW4 forward price is not 1100; it's instead a value between 1101.35 and 1100. There are ways to measure the implied forward from the options market (triangulating through put-call parity) and there are ways to calculate it to some degree of accuracy by taking into account all the index constituents that go ex-dividend between now and the forward date, but I find that a linear interpolation usually works good enough.

    Let me know if this makes sense.
     
    Capt Hobbes likes this.
  6. Tim

    Tim Well-Known Member

    It makes sense, and I agree there is a "forward" component to the pricing.
     
  7. DavidF

    DavidF Well-Known Member

    Gabor Maly likes this.
  8. ACS

    ACS Well-Known Member

    There are two reasons to use a weekly call instead of a monthly to hedge your M3. Because of the higher Deltas, it costs much less cash to buy a 90 Delta call with 10 days or less to expiration than a 56 DTE call and the Gamma of that close call will give improved downside performance if the market drops. The negative is having to roll the call every week but my broker does a very good job of reminding me that I have expiring options in my position. When I roll weekly to weekly I can sometimes do it for flat or a small credit but when I have to roll from a weekly to a monthly it always seems to cost something. I suspect the monthly options have a little extra risk premium because of the AM settlement.
     
    TheSpeculator152 likes this.
  9. Tim

    Tim Well-Known Member

    ACS - I have trouble rolling to the monthly as well. So now the latest I roll to is the week before the monthly. I never stay in the trade past 7 DTE anyway.
     
  10. AKJ

    AKJ Well-Known Member

    I found that looking at the Analyze graph in OV helped me a lot in conceptualizing the pro's and con's of using the weekly v. the regular monthly.

    I set up a position that is long 1 90 delta call with 14 DTE and short 1 90 delta call with 56 DTE. Then I look at the T+0 and T+7 lines. This doesn't consider execution, but certainly helps me visualize and confirms what ACS described.
     
  11. Steve S

    Steve S Well-Known Member

    Actually the formula is C-P = Exp(-rT) * (F - X) so the time premium is P = C - Exp(-rT) * (F - X), and even when rates are tiny the discount factor is significant e.g. for July options we have approximately Exp(-rT) = 0.99924 so the discount is about 0.8 of a RUT point on the strike ... if you are using the theoretical formula for the forward, F = Exp(+rt) * (RUT - PVDiv), then the discount factor cancels out in the forward calculation so the 0.8 points is NOT mostly canceled on the F side ...

    However, the theoretical formulas are not what is really going on, at least for RUT and to a lesser extent SPX ... it's axiomatic that the "real" underlying for a derivative that is not about to expire is the most liquid hedging contract; e.g. for RUT the "real" underlyings are TF/IWM and similarly liquid OTC contracts; for SPX the "real" underlyings are ES/SPY and similarly liquid OTC contracts, where of course there are appropriate basis values added to the hedge contract prices to get the "real" forward RUT/SPX values for each option expiry. For SPX the arb is very tight so ES/SPY will stick to SPX to within about 0.5 SPX points except in a fast market ... but RUT is a very different animal than SPX, arbitrage-wise, and TF/IWM (the "real" underlying price, modulo basis values) gets away from the RUT print very easily even in a mildly moving market.

    Another thing is that, at least recently, RUT forwards have been significantly and systematically divergent from the theoretical PVDiv calculation, which I would guess has something to do with the June reconstitution (anybody have specific expert knowledge of why this is so?) ... I use the PVDiv values from Interactive Brokers; they have always been very accurate for SPX and they seem to be at least ballpark-accurate for RUT (you can tell by calculating the implied dividend yields), but they don't come close to explaining why the RUT forwards are what they are in the market.

    So to definitively answer the question, "is anyone getting DITM RUT calls for free" you have to have the right TF or IWM basis and capture the hedge contract print at the moment you get filled on the call. Personally, I'm VERY certain that it never happens unless you are fortunate enough to cross the trade against retail before a professional grabs it.
     
    Tim R and Capt Hobbes like this.
  12. Steve S

    Steve S Well-Known Member

    Another great discussion in Trading Group 2 about DITM weekly calls ... some questions about exactly why it seems to happen quite often that a call can be rolled for a credit. Granted, it could definitely happen that you just get a great fill because of some big (and/or dumb) trade being done on one side of the spread, but what about a calm market, no big trades pushing things around, and the cash/futures arb is tight? This case is easy to shed some light on (if you want to skip the formulas then the punch line is, it makes perfect sense if the puts are small enough):

    I'll talk SPX since RUT/TF/IWM behavior is so squirrelly that I avoid pretending I understand it ... say C2, P2, T2 D2 = call, put, time and PVDiv for the longer-dated contract and C1, P1, T1, D1 for the shorter-dated contract. Since we're assuming market conditions where the "theory" is actually applicable, we have

    C-P = Exp(-rT) * [F - X] = Exp(-rT) * [Exp(+rt) * (SPX - PVDiv) - X] = (SPX - PVDiv) - Exp(-rT) * X

    Which implies (C2 - C1) = (P2 - P1) - (D2 - D1) - X * [Exp(-rT2) - Exp(-rT1)].

    In today's market with T2 - T1 around 7 days, Exp(-rT2) - Exp(-rT1) is about -0.9999 which is something like -0.2 SPX points, so we have approximately

    (C2 - C1) = (P2 - P1) - (D2 - D1) + 0.2

    But what is a "normal" number for D2 - D1, the difference in PVDivs from week to week? The cuff is that it's currently roughly 1 SPX point per week on average, but dividends on the indexes are famously lumpy so it varies a lot in the closer expirations. For example, using Interactive Broker's PVDivs which are pretty darn good, the PVDivs for SPX expirations for the weekly expirations from 10 days to 59 days are as follows:

    1.86, 3.07, 3.63, 4.01, 5.09, 5.85, 6.33, 6.88

    so the diffs from week to week are

    1.21, 0.56, 0.38, 1.08, 0.76, 0.48, 0.55

    The two key observations are as follows:

    1.
    Just using the average weekly difference in PVdivs of about 1 SPX point, you can see why the call spread could go off at a credit if the puts are small enough: If D2 - D1 = 1 then (C2 - C1) = (P2 - P1) - 0.8, so whenever P2 - P1 < 0.8 the spread SHOULD have a mid price that is a credit.

    2.
    Why would a GTC order for the spread sit there for days and then suddenly get filled? Several reasons, but one good one is that over 1 or 2 days of lumpy PVDiv decay the trade goes from being a mid-price debit trade to a mid-price credit trade.

    My 2 cents.
     
  13. Boomer34

    Boomer34 Well-Known Member

    Great info guys...Thanks for the insight
     
  14. tom

    tom Administrator Staff Member

    Here's the recording of the Trading Group 2 meeting. Enjoy!

     
    TheSpeculator152 likes this.
  15. Steve S

    Steve S Well-Known Member

    More discussion in Trading Group 2 about rolling DITM weekly calls for credits ... this time I jotted down the particulars that were discussed ... here is why the nickel credit received makes sense:

    From the screenshots:

    Strike = 1060
    June10 put value = 0.40
    June03 put value = 0.10

    Using interest rate = 0.5%, X * [Exp(-rT2) - Exp(-rT1)] is very close to -0.10

    From Interactive Brokers, PVDiv(Jun10) - PVDiv(Jun03) = 0.64

    So the fair value of the 1060 call spread is Call(Jun10) - Call(Jun03) = (0.40 - 0.10) - 0.64 + 0.10 = -0.24

    According to this calculation the trade was done at almost 0.20 slippage, which makes sense for a trade like this.
     
    AKJ likes this.
  16. Boomer34

    Boomer34 Well-Known Member

    Thanks for the math...

    I'm trying to grasp the weekly calls...because they make sense to me on multiple levels...as Andrew discussed in the TG2 meeting this week. (Thanks Andrew!)
     
  17. AKJ

    AKJ Well-Known Member

    Thanks for the additional example Steve. Collectively as a community, it does appear that there are still some gaps in the understanding that remain to be filled.
     
  18. ACS

    ACS Well-Known Member

    The biggest credit I've gotten is .15 but .05 is much more typical and sometimes you just have to pay something to get it done. The important thing to me is that by using weekly calls with time value of dimes not Dollars and rolling them on average for flat then it's costing no more and maybe less decay than being long a much more expensive monthly call that performs worse in a big decline where the M3 needs the most help.
     
  19. AKJ

    AKJ Well-Known Member

    ACS, I'm not sure if you have done this (see above quote) in OV and looked at the analyze graph, but this was pretty clear evidence in my opinion of the pro's and con's of the weekly call method. As you mention, the weekly call performs much better if the bottom falls out of the market. It also performs a tiny bit better if the market screams higher. There is cost to this however, and it comes in the form of higher decay if the underlying remains relatively rangebound. The size of this range is dependent on the vols, the underlying, and your choice of strikes, but regardless, there is a fairly wide range of outcomes where the weekly call method performs worse.

    On top of that is the slippage you experience. This method of rolling weekly calls requires making multiple trades on DITM options that are quoted with wide bid/ask spreads and that market makers are unlikely to want. As Steve pointed out, the slippage on such a trade can be quite large and may add up over multiple trades. This is another trade off to consider. It may still be worth it to have the incremental improvement to the downside, but in the end, my gut is that it's negative EV. Similar to buying a DOTM put to protect against the downside.
     
  20. Boomer34

    Boomer34 Well-Known Member

    Andrew...that makes solid sense.

    But, my question is this: Even if the weekly call was slightly more expensive over the course of the trade...the fact that you have more buying power, assuming you use it to buy extra butterflies...way more than makes up for the small costs to roll it and such.

    Now if you don't buy extra calls and utilize your extra buying power...then yes, this line of reasoning should be considered...but if you can go from making on average 5% for a regularly setup M3 to making 7-10% using the weeklies with your increased buying power...isn't that a no-brainer?
     
    Last edited: Jun 2, 2016

Share This Page