Options profit/loss vs probability and the rewards

Discussion in 'Options' started by Melch002, Jul 29, 2021.

  1. Melch002

    Melch002

    I’m new to trading and only paper trade and learn on Thinkorswim. I’m interested in put credit spreads (out of the money) and these trades is what I am testing and learning.

    The following numbers are for the put credit spread of AAPL September 3 2021, with the stock trading at about $ 145. (recorded on July 30 2021)

    The strike 136/135 has a “risk” of 24% expiring in the money and a maximum profit of $ 16 for risking $ 84, this gives a maximum return of 19%.

    The strike 137/136 has a “risk” of 26% expiring in the money and a maximum profit of $ 19 for risking $ 81, this gives a maximum return of 23,4%.

    For 10% increase in risk (24% to 26% risk increase) the market offers 23% more possible returns (the returns increase from 19% to 23,4).

    My question: is this a statistical correct comparison, and if yes, how is this relation/phenomenon between the probability of expiring in the money and the maximum return called? Why is there more than twice as much return compared to the increased risk? I like to learn more about the background.
     
  2. guru

    guru

    I'm assuming that ToS does some custom calculations to show some type of "risk", however they came up with it. Doesn't mean anything since no one can predict the future.
    The market decides what the risk is to the participants, by buyers buying options that seem too cheap to them while sellers selling options that seem too expensive to them. In the end the market decides the odds, and the option prices reflect those odds. Any other calculation of odds is individual to the person estimating those odds. If you have some information that no one else does then you can perform arbitrage by selling options that you know to be overpriced, for example. Likewise, if you think that ToS can predict the future then you can collect money from someone who doesn't have ToS and therefore cannot predict the future. Hmm, sounds like nonsense, maybe because it is :)
     
    .sigma and ValeryN like this.
  3. Atikon

    Atikon

    If you want to papertrade papertrade Interactive Brokers. Tda gives you unrealistic fills
     
  4. guru

    guru

    Btw, I might’ve oversimplified this because people consistently overpay for insurance and therefore selling insurance is a viable business. That’s how many people sell naked puts for income, and selling put spreads can also work. Some funds may also be using options to decrease volatility at the cost of lowering their profit and making it more consistent. So there may be ways to profit by supplying liquidity to those other parties. But there is always some cost, whether accepting higher volatility, or risk of occasional but large losses, etc. I just wouldn’t focus on specific “risk” unless you can estimate it yourself through backtesting and be able to control it and understand it, potentially trading systematically. Otherwise taking some number you see somewhere is meaningless, but if there is a way to prove that the risk is smaller or odds are better than the odds implied by the market, then it may be for reasons like above. While it’s still difficult to beat the market at scale, with some ETFs also trading options but not really doing better than just holding SPY/SPX.
     
    Last edited: Jul 30, 2021
    .sigma likes this.
  5. newwurldmn

    newwurldmn

    You should convert each trade into an expected value: prob of loss*$loss + prob of earning * $gain. And then pick the one that’s higher.
     
  6. Melch002

    Melch002

    Thanks for your reaction. I made the calculation for the AAPL put credit spread with strikes 134/133 (27.5), 135/134 (30,0), 136/135 (32,0), 137/136 (35,1), 138/137 (38,5) and for the at the money 146/145 (51,0), and the closer I get to the current stock price, the higher number I get from your formula. This "feels" not correct, since I'm looking for the highest possibility of keeping the net credit of the sold credit put spread.
     
  7. Melch002

    Melch002

    Thanks. The probability calculations for options are widely used. Do you advise to not use them when deciding that strike price to trade?
     
  8. Melch002

    Melch002

    The "risk" I use is the probability that is calculated for options. When the probability for an option strike price is 80% of expiring out of the money, I assume the probability (risk) of expiring in the money is 20%.
     
  9. Melch002

    Melch002

    In ToS the fill price is de middle middel between Bid/Ask, I think. I'm testing with very liquid high volume options/stocks like AAPL and there is not a wide spread. Do you know the difference with Interactie Brokers? Is their fill more to the Bid or Ask side of the price range?
     
  10. guru

    guru


    Not really. Just the fact that you have to say this is the proof that you provide no substance, no specific formula, no relation to Black-Scholes, and no research papers to point to. That’s like saying “herbs are widely used to cure cancer.”
    AFAIK some commercial company invented this term to make gullible people think that they are looking at something relevant.
    The options pricing is what determines the probability, which means that you can’t come up with different probability that would allow you to profit from discrepancy between option prices. Any discrepancies from known probabilities are arbitraged away by computers that use advanced mathematical models beyond basic probabilities, thus resulting in options prices reflecting realistic probabilities based on all currently known information.
    But to get an idea of probability derived from options, you may use the (single) option’s Delta, which is calculated from option prices and therefore again not based on some other external probabilities. And many people do use Delta to pick option strikes, not always due to probabilities alone, but due to other greeks as well. Some people also look at option price skews that may result from certain mispricing of options, but I don’t think that some unexplained probabilities would give you an edge.
     
    Last edited: Jul 30, 2021
    #10     Jul 30, 2021
    .sigma, qlai and Melch002 like this.