Option historical testing

Discussion in 'Options' started by FLfutures, Mar 15, 2015.




  1. The position gets overrun when the underlying makes an unexpected move. Options are priced so that it doesn't take much for your position to be a winner or loser based on how the underlying moves.

    It's all about the underlying.


    :)
     
    #21     Mar 15, 2015
  2. xandman

    xandman

    The position gets over run when the underlying makes too many moves and you adjust too much.

    I think we have a disconnect between the leverage/optionality vs volatility characteristics. They both exist.

    However, let me say this: Tomorrow, I don't know where AAPL is going, but I see a high probability that certain AAPL options will be priced at higher volatilties for four periods a year than today. (Ofc, I am currently leaving out the current market context.) A known flaw in backtesting. You can actually make money out of that and remain market neutral.
     
    Last edited: Mar 16, 2015
    #22     Mar 15, 2015
  3. xandman

    xandman

    What's noisier? Which would you rather trade?

    upload_2015-3-15_22-44-21.png
     
    #23     Mar 16, 2015
  4. Here's something you may find interesting

    high-quality option data fed to THOMSON
    REUTERS QA DIRECT or other testing platform. Stay away from the retal trash



    http://www.hanweckassoc.com/

    Peace.

    surf
     
    #24     Mar 16, 2015
  5. "Back testing has very (if any) limited use. Show me historical data for stock option strategies from 100-500 stocks over a 1-5 year run and I will show youF*CKING NOISE. Option positions have the same characteristics, in most cases they are VERY CLOSE to being a winner or loser. That's the nature of options, they are volatile and you don't need historical data to tell you that."


    If you're dancing that close to winnners and losers, I would see that as a serious red flag. I think you need to expand your option trading strategies if you really believe this is the case. I assure you, backtesting on option strategies is just as if not more important than in any other trading style. You're clearly a fly by the seat of your pants type of trader, but anything even resembling quant style trading is going to require effective backtesting.

    And also, backtesting is a skill in itself. Designing meaningful tests that avoid curve fitting as much as possible isn't as easy as people think it is. Perhaps your problem isn't the backtest itself, it's how to design them and interpret them?

    I agree with you, the way most people backtest, which is mostly just curve fitting, is nearly useless. But when done right it's just about the most valuable thing you can do when designing new trading systems.
     
    #25     Mar 22, 2015
    N2M likes this.
  6. N2M

    N2M

    [/QUOTE]If you're dancing that close to winnners and losers, I would see that as a serious red flag. I think you need to expand your option trading strategies if you really believe this is the case. I assure you, backtesting on option strategies is just as if not more important than in any other trading style. You're clearly a fly by the seat of your pants type of trader, but anything even resembling quant style trading is going to require effective backtesting.

    And also, backtesting is a skill in itself. Designing meaningful tests that avoid curve fitting as much as possible isn't as easy as people think it is. Perhaps your problem isn't the backtest itself, it's how to design them and interpret them?

    I agree with you, the way most people backtest, which is mostly just curve fitting, is nearly useless. But when done right it's just about the most valuable thing you can do when designing new trading systems.[/QUOTE]


    What ^^^ this guy said.. Day and night if you know what to avoid.. and the backtest you create is YOURS, you did it, you know what went into it.. that is the best confidence builder there is..
     
    #26     Apr 15, 2015