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

Sharing my Road trip backtesting result from 2011 to current.

Discussion in 'General Discussion' started by G..., May 11, 2018.

  1. G...

    G... Member

    Hi guys, I want to share my recent backtesting result on Road trip trade. underlying is RUT, margin is from 6000 to 10000. I'm a very conservative guy, I did backtesting just for gaining some confidence, and I really like the smooth equity curve. I think the only way to minimize the draw-down is to cut the loss early.
    RUT2011_2018.png
     
  2. DGH

    DGH Administrator

    Hi G... Thanks for sharing your results. I don't see your user name among our Road Trip Subscribers, but it was good to see your results with RUT. The equity curve you show is typical for the performance of the RTT...slow and steady increases. The performance of the RTT is also excellent using SPX or ES options, as Tom and I do. I suffered a draw down in the Feb March cycles due to illness (concussion and after-effects) which seriously affected my risk management of my positions. However, as you suggested, the best approach when price passes through the short strike of the core position is to begin to exit the trade. This approach also makes sense on an expectancy basis. Had I been able to effectively manage my risk during the Feb-March cycles, my losses would have been minimal. Tom and I have additional management strategies which are taught to RTT Subscribers. We will soon be moving the RTT to Aeromir.com, but current subscribers will be transitioned easily. For any new traders viewing this thread, Tom and I would like to invite you to the RTT Service. A trial membership is available.
     
  3. FlyingDoc

    FlyingDoc Active Member

    Thanks for sharing these results. i would be very interested in taking a deeper look under the hood. Would you be able to export the ONE data and be willing to share?

    Peter
     
  4. G...

    G... Member

    I'm not RTT Subscriber, all I learned was from free public materials and I'm still learning. My adjustments are slight different, but the trading concept is the same. Here is my summary report exported from ONE.
     

    Attached Files:

  5. status1

    status1 Well-Known Member

    Interesting but I am not sure if anyone can actually make those trades in reality
    From the spreadsheet it looks like a lot of the trades are opened or closed at the exact time of day
    I am not sure if anyone can place and execute a trade at a certain time of day and get it filled at the exact price
    Just curious why there are different max margin amounts ?
    Do these trades have different strike widths ? Do you select the trades or does the software does all the trading
     
  6. Marcas

    Marcas Well-Known Member

    Of course not. What is the problem?
    G... put some work, shared results with community - this is great, I hope more will follow. If he joins Dan's Service he will disappear from public forum with his work - bad. Do not discourage, please. There was no interpretation with graph at all, he even did not say if he thinks results are good or bad (only that he like smoothness of curve) - wise. If we have similar tests for other trades we could compare. Is return 50% per year good or bad (about 3K/y 6K margin)? Depends. Is it realistic? Depends. Is RTT as save as eq curve shows? etc.
    I understand your doubts status. To resolve them more tests is needed. Definitely you can not tell much about trade after doing single run like G.. did. But he did first step.
    G... did you follow fixed rules in your test?
     
  7. FlyingDoc

    FlyingDoc Active Member

    Thans again for sharing. it would probably need a detailed report and significant time to analyse in order to make any indepth comparison to the service generated results however one question did pop out to me from the spreadsheet. What are the exit criteria? I am guessing there is a max loss target or somewhere arund 4-5% as one element but there are several trades that are minimal profit or very small loss with significant DTE left that were exited.

    Peter
     
  8. status1

    status1 Well-Known Member

    I did not mean to spoil all the fun
    I was merely pointing out that while the equity chart looks great but the reality may not be similar

    I am not sure cutting losses early would make sense in all cases unless you know in advance that the trade would not work out
    I am not sure how much it can be learned by closing the trade after one day Perhaps after an adjustment it could have been a winner
    or iit would have worked even without an adjustment if the market recovered
    Closing the trade just to make the equity chart look smoother is not a good reason to close the trade in my opinion
    I did not mean to imply that this test has no value just that it has a lot of unanswered questions which I am not sure it can be answered by merely bactesting
     
  9. Marcas

    Marcas Well-Known Member

    I do not disagree. Your comments were correct to 'typical' BT approach - G... did not expressed this approach. He presented graph without any conclusions.

    Cutting loses in many cases means cutting profits, sometimes quite significantly, so you are right... and the same time you are not :)
    It all depends on your goals. For some traders cutting losses is critical for others it is unnecessary drag on performance. This is why I stressed so much need for individual work, merely following somebody else is not the best option.
    So:
    - all depends on your objectives, if you want smoother eq-line it is one way to get you there.


    To my knowledge there is no class that teaches you how to advance beyond that, and frankly in my opinion such class wont be any good except maybe exposing problem . I would like to see exchange of experience between interested traders that is why I asked not to discourage any such attempts.
     
  10. G...

    G... Member

    I can share the details of my backtesting data. My backesting of course is not perfect, but at lease it gave me some rough ideas on how good the trading system could be. I don't set profit target by a singe trade, I set my profit target by months and years. I exit the trade base on various reasons, such as quick profits pop up, too directional, less profit potential left or etc. I reentered the trade after that. I don't view it as "exit the trade ", I view it as "reposition" . I don't fight downside, I don't do baby butter, I like to keep my trade as simple as possible.
     

    Attached Files:

  11. Marcas

    Marcas Well-Known Member

    I think it is all good, in my opinion you are doing it fine to get quick overview.
    I wanted to code in your plan but with above it will take too much time, especially that I do not trade this. Sorry, I'm on smth else at the moment.
     
  12. FlyingDoc

    FlyingDoc Active Member

    Thanks for the detail export G. I have imported these trades in to my ONE and hope to find some time to review soon. Unfortunately the ONE import tool i have is not great at seperating the trades. I grouped everything by expiry which results in fewer total trades than your performance graph indicates. i may be able to manually seperate these trades as i review them. I agree with Marcas that if the 'rules' for adjusting the trade are not clearly defined then its harder for anyone else to evaluate or recreate the testing.

    Thanks again for sharing
    Peter
     
  13. garyw

    garyw Well-Known Member

    @ FlyingDoc: It appears G... 's report breaks a "reposition" into separate trades. Unclear if this should negatively impact the usefulness, as the total PnL should match, as a validation.
    Seems that breaking by expiration (as you have done) may actually be more ideal (more like the way the RTT is normally traded).
     
  14. FlyingDoc

    FlyingDoc Active Member

    i agree Gary. I was just highlighting that using the ONE analyze reports will generate different summaries and statistics... the net profit of all the trades is indeed equal.
     
  15. Mark17

    Mark17 Well-Known Member

    I wanted to pick up with status1 and Marcus' discussion because I think it is a very important one that is seldom had.

    I totally agree about the importance of knowing/understanding the rules used to properly analyze a backtest. Many potential flaws abound when it comes to backtesting and multiple sets of eyes to evaluate can help to detect/avoid these potential flaws. This does not usually happen, though. What usually happens is that someone throws up a backtested equity curve or table of results that is seldom questioned. I think part of this is because most people have little education about trading system development and do not understand many of the finer details.

    I don't see any discussion of rules used in G's backtest here. To me, this means the backtest is suitable to G. If it gives G the confidence to trade and stick with this system then great: that is one of the best reasons and benefits to do the backtest in the first place. For the rest of us who don't know the nuances of G's backtest, though, it means very little. I wouldn't trust my money with a strategy I did not understand completely. Would you?
     
  16. Marcas

    Marcas Well-Known Member

    This is only one of many possibilities. If true, next question is: why is that.

    He did not post any rules he went by.

    I with you about importance of topic, just have to recognize that not everybody agree with us (or care).
     
    Mark17 likes this.
  17. status1

    status1 Well-Known Member

    I also agre with back testing but I am not sure how valuable information this would be to other traders here
    I am not saying not to share the back test results I just would have preferred to see real results that can be discussed among traders
    To me this is like a seed that has the potential to grow into something big or not grow at all but until the seed is planted or the trade is made this is all hypothetical
    Having a smooth equity curve is nice if it can be achieved but until this is traded for real it's just something to look at
    I would have preferred to see real results even if it's not smooth This way it would show all the flaws and perhaps show where a trade could have been managed better to avoid some of the bigger losses
     
    Mark17 likes this.
  18. Mark17

    Mark17 Well-Known Member

    Multiple possible answers. I think many educators know little about it. I think a background in statistics, which many people don't have, can help a lot. I also think greed plays a role: people see results that impress them and simply want to learn the trade in hopes of making future gains. This is more fun than dissecting the approach and its validity (including ample sample sizes)--regardless of whether they believe in a need to do so.

    I definitely agree that many people don't care.
     
  19. TraderJ

    TraderJ Member

    It could also mean the test is curve fit or has "look ahead" bias which would invalidate the results. The best way to back test is to formulate a set of rules and test them on historical data, then do another "out of sample" test on different data. Even people who do very rigorous/careful backtesting design systems that work well in the past then totally flop in real time. I've seen that happen many times--enough to make me very skeptical.
     
    Marcas, Mark17 and status1 like this.
  20. Mark17

    Mark17 Well-Known Member

    That's a great reason why I think collaborating with others is far superior to one person backtesting alone. He may unknowingly have a completely flawed design if others did not have the opportunity to proof it.
     
    Marcas likes this.

Share This Page