MultiCharts Automated trading

[Bad_Badness]

Have you been able to quantify what the slippage, poor fills, and latency has been costing you ?

No, but it is doable, but not worth time to get accurate.

For the workbench-prototype-model as expressed in EL, I use a generous fixed padding: "through 1 tick" fills + 2 ticks for slippage+comms.

I expect if you take something that is a MC workbench version and try to actually trade, you realize you need a C version eventually, if not immediately.

I think where you can waste time is trying to duplicate or overly workaround the interface's (MC) order engine. You build your own order handling-router, even if it is primitive. And you can prototype a primitive order handling in EL's logic, somewhat.. You can define your custom Price/Time function.

BTW: there is a bunch of stuff on the MC forum about their order engine.

 

[MarkBrown]

MultiCharts will not autotrade as advertised. I will pay anyone who can give me sample code that actually matches backtesting with real trading.

 

[TrAndy2022]

MultiCharts will not autotrade as advertised. I will pay anyone who can give me sample code that actually matches backtesting with real trading.

Oops, I found a solution some years ago. It is possible but some efforts are needed. You also could ask user Smoky at MC forum who has a solution found. It is possible. PM me if you want.

 

[VSTscalper]

MultiCharts will not autotrade as advertised. I will pay anyone who can give me sample code that actually matches backtesting with real trading.

I have used Tradestation - NinjaTrader - and Multicharts - am not impressed with any of them as far as going Full Auto. Currently I use Multicharts only - I just do FAST Scalps - keep things simple. For me - it is all about making my Net Daily Goal - within 2 hours of the Open - then done for the day.

As far as Backtesting goes - personally - I find it to be almost Total BS - the actual results never seem to be as good as the Backtested results. With my Code - while I can Backtest - I do NOT need too.

The way I code - I am only concerned about what is happening on the Current Bar ONLY- not 2 days or bars ago - not 100 days or bars ago or longer. See the snapshot of how far back I look.

I am NOT saying that my Code is perfect - just saying - it works great for me. I went thru the process of Coding - using Indicators - for many years - with mixed results. Then - I started doing some Meetups with guys from Microsoft and Google. They are programmers - AI and ML - this totally changed the way I Code. By the way - I live in the Seattle area - East side of Lake Washington - and between the Main Microsoft Campus - and Google Campus. It has been fun picking their brains - and believe it or not - many of the programmers - do Trading on the side.

Was just looking - I had sent you a Private message - close to 2 years ago. If you want to talk sometime - I will send you my Skype name.

 

[syswizard]

MultiCharts will not autotrade as advertised. I will pay anyone who can give me sample code that actually matches backtesting with real trading.

If you use any of the Set functions (SetProfitTarget, SetTrailingStop, etc) or have [IntrabarOrderGeneration=True] in your code, it will NEVER be able to achieve backtest=realtime results.
One workaround is to create functions that mimic the SET functions. For instance I created my own SetPercentTrailing function that only operates at the end of the bar.

 

[MarkBrown]

so it seems that MultiCharts has at least given the user some more choices how to form range bars.

using 1 tick resolution to make kase range bars is the only combination that will not give phantom bars and or phantom price ticks. what you see is what you get, this is huge, but still running in sim mode.

however i have been watching some systems taking trades in real time and at least i know i could have traded them also manually with the same results.

the next step is to flip the thing on to a live account and see what it does, i will go slow on micro first lol.

m

ps but at least i am making some hopeful progress, which i had just about given up with automated trading on MC.

 

[syswizard]

the next step is to flip the thing on to a live account and see what it does, i will go slow on micro first lol.
ps but at least i am making some hopeful progress, which i had just about given up with automated trading on MC.

Don't give up Mark....just be sure your internet connection remains fast and you have elevated the Windows priorities of MC.

 

[VSTscalper]

so it seems that MultiCharts has at least given the user some more choices how to form range bars.

using 1 tick resolution to make kase range bars is the only combination that will not give phantom bars and or phantom price ticks. what you see is what you get, this is huge, but still running in sim mode.

however i have been watching some systems taking trades in real time and at least i know i could have traded them also manually with the same results.

the next step is to flip the thing on to a live account and see what it does, i will go slow on micro first lol.

m

ps but at least i am making some hopeful progress, which i had just about given up with automated trading on MC.

Personally - I really prefer trading Manually - try to reach my Net Daily Goal in Dollars - within the First 2 hours of trading.

I do use Range Bars - instead of Kase Bars - one reason being - I use the Bar Magnifier to look Inside the Bar. The Bar Magnifier is NOT available with Kase Bars.

As I have mentioned previously - I look - Inside - the Current Bar - ONLY - to Enter a trade.

Based on this snapshot below - under - Properties - Backtesting - the Kase Bar is considered ------ a "non-regular chart".

Whatever works for a trader - that is what they should use. I finally I stopped looking for a program that worked consistently trading Full Auto. Even on Full Auto - I would still be watching the trades - because there are things that can happen - totally out of our control - that can cause Full Auto to - FAIL.

 

[MarkBrown]

Personally - I really prefer trading Manually - try to reach my Net Daily Goal in Dollars - within the First 2 hours of trading.

I do use Range Bars - instead of Kase Bars - one reason being - I use the Bar Magnifier to look Inside the Bar. The Bar Magnifier is NOT available with Kase Bars.

As I have mentioned previously - I look - Inside - the Current Bar - ONLY - to Enter a trade.

Based on this snapshot below - under - Properties - Backtesting - the Kase Bar is considered ------ a "non-regular chart".

thanks for the input it helps talking to traders doing similar things.

classic range bars 4-32 tick range (or points as mc calls them)

mostly i trade on the close of the range bar or open of the next range bar and that's how i test also so i don't really have a need for the inside bar look. but on classic regular range bars i have seen a bar form and place an order only to have that price tick it traded disappear, leaving my entry arrow with no bar showing where it traded.​

standard range bars 4-32 tick range (or points as mc calls them)

using the standard range bars, like others have stated the gaps will be filled in with fake price bars.​

kase range bars set to 1 tick update resolution 4-32 tick range (or points as mc calls them)

this combination gives no fake bars and no repainted price bars, at least i have not observed any yet.
​

net daily goal

i like to trade 8:30am to 9:30am and maybe from 2:00pm to 3:00pm chicago time

bar magnifier

just wondering what size bars you using for that?

 

[syswizard]

Nice summary Mark. I don't quite understand how the gaps occur.