Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
What is in your opinion the best order platform, with that i mean, i have a stand-alone bot, it takes in data-feed, does all the calculations and generates and manages the trades.
I build this stand-alone, to have structured code, compared to many different indicators (this has a significant development price). But i control it from end-2-end, which is an advantage.
Today i inject orders into NinjaTrader, but the ATI api is kind of limited in my opinion, if you touch the orders in chart trader, through the API you are blind. I can work around going to the sdf database, but that is not officially supported. The hole setup is also not 'rock solid' as NinjaTrader has it's own character and crashes from time to time especially if the markets are busy and you 'poke' to frequent through the ATI. The nice thing is that i can connect to a big series of brokers without extra effort.
Now my question, is there a better solution for broker abstraction, that allows better control over the order life cycle ?
The level of depth you seek is out of my league - way beyond my knowledge base. I have been running Tradestation automated for years, and never really searched for an alternative (I have not seen a pressing need to). I have a pretty solid backup plan in place if Tradestation brokerage and/or platform suddenly goes away, but it is nowhere near what you have done.
Trading: Primarily Energy but also a little Equities, Fixed Income, Metals and Crypto.
Frequency: Many times daily
Duration: Never
Posts: 5,057 since Dec 2013
Thanks Given: 4,409
Thanks Received: 10,225
@rleplae Have you looked at Trading Technologies XTAPI? I have no personal experience but I gather it's top notch. Not as cheap as NT or Tradestation though.
I use walkforward optimization on almost all my strategies. I usually use rolling forward walkforward. For certain fitness functions, anchored WF is more appropriate.
Are the PhDs traders?
Walkforward can be curve fit, if you do it incorrectly.
I understand that you use Stratopt for WFO and WFP. Have you used Diamond Backtesting with Walk Forward Manager (BTWFMgr) by profsoftware? It will be interesting if you can create another video comparing Stratopt and BTWFMgr.
Unanchored wf might not be correct, if you use a fitness function like drawdown - depending how your walkforward software calculates it. It is hard to explain, since it depends on the particular software. Example why this MIGHT matter:
Date1: Jan 1, 2000 Net Profit = XXXNP, Max drawdown= XXXDD
Date2: Jan 1, 2010 Net Profit = YYYNP, Max drawdown= YYYDD
Net Profit, from Date1 to Date2 = YYYNP - XXXNP
Max Drawdown though is not YYYDD-XXXDD
Curve Fitting Walkforward
2 examples:
A. You run walkforward, get results and then think "I can improve that with a filter" and you do this say 4 or 5 times. Everything improves and you know you've made out of sample walkforward results better. Problem is your walkforward is not really out of sample anymore - it is some conglomeration of out of sample and curve fitted results.
B. You pick 10 combinations of IN/OUT to do your walkforward. You pick the best one to use going forward.