Backtesting is useless

Discussion in 'Options' started by JCDST1979, Jun 30, 2020.

  1. Heydrrich

    Heydrrich

    Well personally i'd never say that I know everything and I assumed most people would think this. It seems that I have learned that there ARE really smart people who know it ALL and yet visit ET :D:D:D .....

    =============================

    Anyway, I am asking these questions because I'd like to define the problem of backtesting, and understand what others see as backtesting, in order to determine whether or not it can be useful. So I am trying to learn here.

    H
     
    Last edited: Jul 6, 2020
    #91     Jul 6, 2020
  2. virtusa

    virtusa

    If these smart people think they know everything, they only proof not to be smart. In general stupid people think they are smart.

    You don't need to know everything (how can you even know you know everything?), if you are smart you understand that even with a decent amount of knowledge you can survive in trading.

    I don't know much about trading, but I understand the basics and master them well enough to be consistently profitable.
     
    #92     Jul 6, 2020
  3. ironchef

    ironchef

    But according to @Buy1Sell2 it is the Only True Edge:

    https://www.elitetrader.com/et/thre...ment-is-the-only-true-edge-in-trading.292728/

    What if risk management can turn a negative expectation method into one with positive expectation?
     
    #93     Jul 6, 2020
  4. Tradex

    Tradex

    Again, the question is very simple: a backtest shows that trading system X has been losing 10% a year on average, for the last 30 years.

    Would you trade such a system, now?

    Would you buy this system, if it was for sale?

    A simple YES or NO will do.
     
    #94     Jul 6, 2020
  5. ironchef

    ironchef

    Just come to ET and you will meet all the professionals and an occasional mom & pop amateur retails like me.
     
    #95     Jul 6, 2020
  6. ironchef

    ironchef

    Absolutely. I would.

    Trade the inverse of system X.

    I am not kidding, I backtested buying put options on SPY, going back 25 years and consistently got negative returns. So, I don't buy puts, rather I sell puts.
     
    #96     Jul 6, 2020
    Tradex likes this.
  7. Prepare to be baffled! :)

    If a trading system loses by doing buys, doing the reverse (i.e. selling) IS NOT going to win!

    Why? Because of the spread. Think the fair value in terms of probability was 50%, with bids at 45% and asks as 55%. If you buy @ ask, you'll lose 5% on average. If you do the reverse, guess what? You'll sell at bid, still a 5% loss on average :)

    Crossing the spread and still making money, that's hard.
     
    #97     Jul 6, 2020
    d08, Heydrrich and rb7 like this.
  8. Tradex

    Tradex


    Losing system X trades once a year. Here are the results for the last 5 years (in US dollars):

    2020 (so far): -75,000
    2019: -45,000
    2018: -125,000
    2017: -85,000
    2016: -15,000

    Total loss: $345,000

    Question: how much money would you have now if you had simply bet against System X?
     
    Last edited: Jul 6, 2020
    #98     Jul 6, 2020
    ironchef likes this.
  9. ironchef

    ironchef

    I tested writing puts instead of buying, they were consistently profitable. True, I didn't account for the spread. Also, I could have bugs and wrong assumptions in my codes since no one checked them for errors.

    In real life trading I actually did something similar. When I first started, I mechanically wrote covered calls and didn't make any money, so I reversed and wrote cash secured puts, came out better.

    Anyway, I am not defending what I did, just wanted to present the other side as food for thought.

    Take care.
     
    Last edited: Jul 6, 2020
    #99     Jul 6, 2020
    Tradex likes this.
  10. ironchef

    ironchef

    Options 101: Never buy on ask and sell on bid. Always use limit price to trade.
     
    #100     Jul 6, 2020