Making a binary bet with vanilla instruments

Discussion in 'Options' started by CFerret, Dec 1, 2019.

  1. Hello,

    Say I have an opinion that instrument X is above/below price Y at time Z with probability of say 65%. How could I make a binary bet on this outcome using vanilla instruments on major regulated exchanges, such as CME?

    Appreciate any advice!
  2. guru


    I believe you’d simply buy a basic call spread or put spread, aka vertical spread.
    In fact, the pricing of binary options seems equivalent to vertical spreads on vanilla options, they just make strikes and expiration times different to confuse people and make it difficult to figure out they can do the same using vanilla options..
    The probability is priced into the options, so you’d just need to make sure your probability is high enough to justify the cost of the spread you’re buying. Say you can try getting $10 wide spread for $5 (usually near ATM), targeting the price you believe to have 65% probability of being exceeded.
    Last edited: Dec 1, 2019
    CFerret likes this.
  3. Epicurus


    Use Bull Put/Bear Call credit spreads.
    I'm papertrading this atm to optimise strike selection.
    Just comes down to whether you can genuinely pick a binary outcome 65% of the time.
    tommcginnis and CFerret like this.
  4. I've looked at this some. And like others have said it's easy to construct a 50/50 R/R with a simple vertical credit or debit spread. Almost seems like a no-brainer: even money bet, with a 65% win rate. You'd own Vegas.

    The catch is your time frame. If your underlying move is not near the vert expiry, you are not going to make that much, unless of course it zooms past your strikes. And to hold it to near expiration invites a reversal. If my systems indicate a move I just grab a near money long option. Limited loss, "unlimited" gain, and no short option hassles. :)
    CFerret and guru like this.
  5. bln


    Have used vertical debit spreads to setup 1:1 binary trade plays.

    I found credit spreads is no good as they usually does not give you an 1:1 ratio but more like 1.5:1 or 2:1, ie risk side is heavier than the profit side.
  6. ironchef


    Here is one binary option trades at NYSE called BMY RT, which was formerly CELG CVR:
    You either get $9 or $0, after March 31, 2021.
  7. Epicurus


    Time frame is what your analysis tells you is required to support your target win rate. I trade Bull Put/Bear Call credit spreads on 4 to 7 wk duration.

    Depends on how close to or even in the money you set the written leg, you can get between 50% to 100% return on capital per trade. I'm testing moving up from 60% to 80% ROC at entry, which being more aggresive means your win rate and/or profit % on close dials back a bit. Need to optimise.

    Winning trades you can let expire to save the brokerage. Losing trades usually will become clear part way through and you can close to reduce average loss.

    It seems to work very well for what the OP is wishing to do. I target a 60% win rate and getting a positive expectancy so far on a small number of trades. Not enough data yet to prove the approach but the methodology is sound for a series of binary bets.
  8. Yes I agree bull/bear spreads can be a good way to trade, but to me this is a special case. In this you hope to keep max win somewhere near max loss. So the way I think of these 50/50 trades is more like a coin flip, but a coin that lets you win around 65% of the time.

    Here's one: You expect QQQ (about 202.7 now) to drop by Friday. Your system says you win 60+% of the time. You sell QQQ 12/6 203 C, buy 12/6 203.5 for about 0.25 credit. So you win or lose $25 with almost any move. But that's at expiration. Say it moves down to 201 by Wednesday and you close it. You are only going to get about $11. Even if you don't take a lot of max losses, taking very many less than max gains will wreck the whole 50/50 thing. I know you and most other option traders here knows this stuff, so this just for newer traders in case they were thinking of trying something like this.

    Last edited: Dec 2, 2019
  9. Epicurus


    Optack, whats wrong with a coin flip with 65% win rate? I think you need to reflect on your numbers a bit more.

    Firstly, you have the stock at 202.7. If you write a 203.00C/203.50C credit spread then at expiry you get 100% of the premium you received (less brokerage). Why close early? The stock can go where-ever it likes during the trade. You take your profit at expiry.

    The trick is picking a series of trades with a 65% win rate. At expiry. You can either pick these at 65% win or you can't.

    Secondly, you then have an opportunity to improve returns. On my timeframes of 4 to 7 wks, say average 6weeks. I'll have 3 weeks or so where these spreads will still hold a fair bit of value to get a look at the trade. By closing just those subset of the 35% losers that are clearly not going to recover, you could reduce your average losing trades to something below 100% .

    This is a very profitable and very simple trading approach, almost set and forget, IF you can pick whether a stock will finish above or below the current price in the timeframe that YOU set at a sufficient win rate.
  10. Yes if I had the coin (system) that would win 65+% of the time, at a certain date, I'd be set for life. I have lots of systems that win greater than 65%, but picking when that happens (like at expiry) is tougher. Especially on shorter time frames. For example, you might think SPY, after all the long bull markets, would have a higher percent chance of being up week to week. But my data shows only a little over 50% probability:

    Here's 20 years of spy, Friday to Friday:

    TOTAL WEEKS = 969
    DOWN WEEKS = 427
    UP WEEKS = 542
    PCT UP WEEKS = 55.93
    PCT DOWN WEEKS = 44.07
    AVG PCT UP = 1.60
    AVG PCT DN = -1.78

    UP CHNGS >= 3% = 69
    CHANGES BETWEEN -2%/+2% 683


    ONE IRUP = 114
    TWO IRUP = 51
    THREE IRUP = 35
    FOUR IRUP = 15
    FIVE IRUP = 17
    SIX IRUP = 6
    SEVEN IRUP = 2
    EIGHT IRUP = 2
    NINE IRUP = 1
    TEN IRUP = 0


    ONE IRDWN = 136
    TWO IRDWN = 64
    THREE IRDWN = 25
    FOUR IRDWN = 14
    FIVE IRDWN = 1
    SIX IRDWN = 2
    NINE IRDWN = 0
    TEN IRDWN = 0
    Last edited: Dec 3, 2019
    #10     Dec 3, 2019
    Epicurus likes this.