Forex Robots - legit or scam?

[MS+Tradesim]

if u dont test youll end up with a divide zero error or something which will wipe out your trading account in 5 minutes.

not wanting to be alarmist.

Oh, I get that and I test religiously...just not using a dedicated test module the way wabbit describes. However, after reading his post a few more times I now get what he's saying and could certainly come up with something to achieve that (given enough cups of coffee and nurofen).

 

[pilbara]

initial testing of version 2 of the derwent (named after the derwent river - which wasnt dammed and was allowed to 'go with the flow' ... ie the trend).

for "no dams" maybe the Franklin would be a better river than the Derwent!

 

[Stormin_Norman]

for "no dams" maybe the Franklin would be a better river than the Derwent!

am i getting my tasmanian rivers messed up?

 

[pilbara]

the Franklin river is famous for the "no dams" campaign in the early 1980s when the Tasmanian State government were trying to build further dams in the South West Wildnerness. Green activists led by Bob Brown blockaded the site and organized a national campaign with major street rallies, until 1983 when the Hawke Labor government was elected and took the Tasmanians to the high court to stop the project.

 

[Stormin_Norman]

its going to have to be renamed then!

 

[pilbara]

yeah the Derwent river is more of an estuary where the tide controls the flow, so maybe that's a better name for a different type of EA.

At the moment I'm working on a trend-following EA which is a less frequent trader than a scalper. I'm trying to get simple parameters on the strength of trend and the time scale. I think it will be possible to have a number of these trend followers running at the same time on the same instrument but using different time scales, eg hourly trend, daily trend, weekly trend, monthly trend.

 

[Stormin_Norman]

thank you for correcting my poor geography.

 

[Punisher]

Oh yea, I'm feeling it. Crapping my paints here. But I'll let it go. I can make it back again if it runs out of free margin tonight.

Its possible I'll lose up to $200.00 USD, but well see.

EM, if you reduced the amount of possible open trades at one time to preserve your capital until your trading account builds up to a reasonable amount? That would reduce the risk of margin calls and also it could save you from filling your pants with crap to.:twak:

 

[MS+Tradesim]

To coin a word, my cellbot (Excel driven autotrader) is now up and running on demo. I seem to have caught and corrected the majority of bugs and bad logic, but I'm sure more will present themselves so will be testing for awhile.

One thing that I really like about this setup is that through IB I do not have to trade lots. I can position size precisely based on risk. For instance I have orders in now for 166481 on USD.JPY and 100537 on EUR.JPY. They update dynamically based on the changing AUD rates and change in entry point.

 

[Stormin_Norman]

that's a very clever system, to use excel to trade.

i am not really sure how that's done, and the instructions passed over, but very clever and well done!

good to hear its operating well after some modifications. are u trading live demo?

 

[pilbara]

... through IB I do not have to trade lots ...

I thought that you had to trade multiples of the minimum order size for it to be routed through IdealPro, otherwise it gets put through Ideal? For example if your account is based in USD then the minimum is $25,000.

 

[MS+Tradesim]

Hey stormin,

I'm not sure what you mean by "live demo" but the IB Paper Trader account uses live data and 'executes' orders realistically such as partially filling (if not enough on offer) or running through as many levels as necessary to fill (on market orders), or getting swept along with slippage, etc so it is probably as real as can be on a demo.

As for my overall autotrading the basic setup is quite inelegant and cumbersome but looks like it will do the job until I can learn C and write my own interface. I've heard that using DDE through excel is not as fast and robust as a program written for the task.

IB feeds the FX data into Metastock and FX/account/order data into Excel. I have an EA (doesn't work the same as MT4 EAs) setup in Metastock that sends entry and exit signals and prices to a database. Excel then grabs that data and calculates position sizes based on %risk, equity, etc and then sends/cancels/modifies orders by placing them into IB. I actually think Amibroker would be able to achieve all that way more efficiently, but screw it! I love MS.

And because Excel is in a real sense, the front end, the next step which will be easy to do is to log all trades and extract the results for assessment, tax records, etc.

 

[MS+Tradesim]

I thought that you had to trade multiples of the minimum order size for it to be routed through IdealPro, otherwise it gets put through Ideal? For example if your account is based in USD then the minimum is $25,000.

There are minimums but beyond that it's still working whether it routes into IDEAL or IDEALPRO.

It's been awhile since I traded FX live on IDEALPRO and I don't remember right now, but if it turns out it does work in lot sizes that it will be easy to allow for and adjust - but annoying from a position sizing perspective.

 

[Stormin_Norman]

Hey stormin,

I'm not sure what you mean by "live demo" but the IB Paper Trader account uses live data and 'executes' orders realistically such as partially filling (if not enough on offer) or running through as many levels as necessary to fill (on market orders), or getting swept along with slippage, etc so it is probably as real as can be on a demo.

As for my overall autotrading the basic setup is quite inelegant and cumbersome but looks like it will do the job until I can learn C and write my own interface. I've heard that using DDE through excel is not as fast and robust as a program written for the task.

IB feeds the FX data into Metastock and FX/account/order data into Excel. I have an EA (doesn't work the same as MT4 EAs) setup in Metastock that sends entry and exit signals and prices to a database. Excel then grabs that data and calculates position sizes based on %risk, equity, etc and then sends/cancels/modifies orders by placing them into IB. I actually think Amibroker would be able to achieve all that way more efficiently, but screw it! I love MS.

And because Excel is in a real sense, the front end, the next step which will be easy to do is to log all trades and extract the results for assessment, tax records, etc.

nice work, very impressive. you must be quite a excel wizz.

by live demo i mean paper trading 'as real' just substituting pretend money for real money.

 

[pilbara]

As for my overall autotrading the basic setup is quite inelegant and cumbersome but looks like it will do the job until I can learn C and write my own interface. I've heard that using DDE through excel is not as fast and robust as a program written for the task.

I think TWS is java, so the java API might be the easiest to work with. Here's a tutorial:

http://www.stockbotprogramming.com/javaibtutorial1.php

 

[Stormin_Norman]

we're looking at preferably going to java when we need to upgrade our broker.

the programmer likes that most out of all languages.

 

[MS+Tradesim]

nice work, very impressive. you must be quite a excel wizz.

Not at all. Just learning as I go. Google is my best friend.

 

[Stormin_Norman]

google can make anyone an expert

 

[MS+Tradesim]

Yay! Found a looping error which keeps placing orders. I think there were something like 50 orders sent.

 

[Stormin_Norman]

can i suggest u start your own thread to keep us up to date with how youre going?