Jump to content
  • 0
Sign in to follow this  

PRT Suggestions - backtesting and automation enhancement

Question

Hi All,

 

Some additional features that would be a welcome addition to PRT...

 

1) Walkforward testing. Problem with backtesting strategies is the optimization of variables specific to the testing period. Without the ability to walkforward (test a period for which the strategy was not optimised), it is easy to create systems that tests well, but trade poorly in real life. Though it is possible to do walkforwarding manually, it is very time consuming and error prone.

 

Most decent testing platforms support this feature.

 

2) Look-Inside-Bar Back-testing.  Currently, if you have a stop and a limit, PRT does not have the ability to determine which is triggered first, if both are triggered within the same bar when backtesting. LIB solves this problem. Platforms such as Tradestation provides this type of functionality.

 

3) Disabling variables. This is just annoying and a huge time waster. Before you can enable a system for auto trading, you have to remove all the variables that you created when backtesting. But then, when you decide to do some more testing, you have to recreate all the variables again from scratch; including name, min, max, and step value. Do this a couple of times and you realise this is not the best way.

 

4) Margins. Used to be there, and now is gone. %Gain (and capital needed to trade a system) is not correct if margin is not taken into account.

 

I've posted previously about some other enhancements that I believe are currently lacking - such as being able to create test templates; where the initial capital, date ranges, spreads, commissions, etc. can be specified once and then used for testing multiple strategies, etc.

 

I would appreciate some comments and additional suggestions.

 

Regards

Stef

Share this post


Link to post

2 answers to this question

Recommended Posts

  • 0

Number 2 is important, and maybe some kind of warning should be in place when starting the backtest?

 

about number 1 i am not sure if this helps but you can choose date and time when optimize a backtest, you can optimize 2015-01-01 - 2016-01-01 and then run the backtest from 2016-01-01 to 2016-03-17

 

about number 3 you can duplicate the system with variables and then only change the system you run in proorder

Share this post


Link to post
  • 0

.

 

2 - That is exactly how I do the walkforwards currently. Would just be so much quicker and easier if this was a feature of the product.
3 - True. Again, I feel this should be a feature. If you have to duplicate each system you end up needing to maintain both sets. Over time they can easily get out of sync.

Share this post


Link to post

Join the conversation

You are posting as a guest. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
You are posting as a guest. If you have an account, please sign in.
Answer this question...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Sign in to follow this  

  • Member Statistics

    • Total Topics
      12,422
    • Total Posts
      63,139
    • Total Members
      83,826
    Newest Member
    Oadebayo
    Joined 19/09/20 05:01
  • Posts

    • I want to believe that, even if we had viruses for many years and we lived with them like nothing happened, this one is killing more people than it should. At least is what reports are saying. That is why countries are forcing lock-downs. Think about that a lock-down is going to extend the furlough policies all over the world again. I don't think a government want to pay that money out to citizens if they don't have to, and companies are going to have to close business for good if we have a second lock-down everywhere. I agree with the statement that we had many corona-viruses before i.e. common flu, but why now governments are taking these draconian measures to force people into a quarantine? If we had the same dangerous viruses always around why are they now locking people at their homes. Did they not care about people before? I think the virus is way more deadly than they tell us and that is why people are not really that scared of it. If we had all the information that is available to professionals we would be really scared and people would lock themselves without waiting any government to tell them. My wife is Chinese and I get their news first hand. They are not reporting what is really happening since the very beginning and it is really hard to trust the Chinese government  on anything related to this virus. They still have the problem there, they keep locking entire cities down for short periods the minute they see anything. It is not reported here in the news but every week they lock people down at their homes.  This virus is really nasty and if locking people down is going to stop it, let's do it. 
    • I agree with 99% of this - the current strain is virtually exact to a strain found in 2012 by a group of Chinese scientists, then when 6 of them died from exposure the virus was sealed and sent to Wuhan I challenged David Icke  to prove the excess deaths were'nt excess deaths, from the info I looked into I'm satisfied the virus is genuine and I'm probably one of the worlds biggest sceptics  From a lock down point of view I'd of thought the perfect place to test out compliance of mass herds would be the USA - hasn't happened strongly enough for me to be concerned but it shown those that could create such circumstances how compliment most people are.  
    • Corona virus has been with us since forever we just never mass tested before. There are no extra deaths this year than any other and it's the same unfortunate types who die of corona virus every year (over 80's with pre-existing conditions 90% of cases). Corona virus is always circulating in the population most people being completely unaware and asymptomatic, herd immunity is built and the virus mutates or dies out. Tests just show you have been exposed to the virus at some stage which could have been months ago and you were most likely asymptomatic. Vaccines don't work for corona virus because they mutate very quickly which is why there is no vaccine for the common cold (a corona virus). And why flu shots don't work very often because they are based on the previous years dominant strain not the current year's (the mass manufacturing would be too late). Covid 19 has already been mutating and the last studies I saw several weeks ago confirmed that as usual a particularly nasty strain becomes less deadly with each new mutation. Another 30,000 people will die next year from respiratory infections as it's the 5 year rolling average (covid 19 or not), weather it'll be covid 19 next year or another strain no one knows but they will keep the covid ball rolling anyway. So all in all it looks like lockdown til the end of time.      
×
×