Hi
I have currently developed a custom indicator in eSignal which appears to be working well. I want to try and test it out on a few different instruments and wondered what would be the best software. Ideally I would want to use the same software to test the historical P&L, optimise it, and put it into production.
There is a strategy functionality in eSignal but it appears to be not based on tick data so I am very wary that it would not give accurate results. When I tried to build a strategy in it a few years ago I found that there was poor documentation, but I got something which might have been working but it was hard to verify. I am not great in JavaScript but am getting a bit better. I also a bit wary of using signal as a trading package - as it has its origins a charting packing but am not sure if this is justified (perhaps this is is due to bad experiences with CQG integrated client when it came out).
I also have an open e cry account (which I don't use much but have a free demo as I have left a little bit in their account) which has real tick data. When I was mucking around a few years ago I have worked out how to code some similar indicators in C#. These would be easy to adapt and in theory to create a strategy to test. But it was really frustrating to code in a language which I am not familiar with. OEC can also autotrade through into Excel through DDE which I understand is old technology and best avoided if I am starting from scratch. (I am most comfortable in VBA).
Does anybody know if OEC can see the code that I would be using for my autotrading / indicator (As they are the broker they can see who are the successful clients - or am I being paranoid?) Other negatives are that I am limited to just one broker.
I also have tried Ninja Trader in the past but this couldn't handle exchange based spreads and I couldn't work out how to export the data to check that my code was correct using Excel.
For my trading at the moment, I use Easyscreen which can autotrade through RTD Excel but it is expensive (I was using it for the autospreader which I probably won't need if this model works as I can get away from synthetic spreads). Although this is robust and properly supported - it won't help me with the testing and optimisation and would only really work for putting the model into production. If I went down this route I would have to buy the tick data and analyse it in Excel / VBA.
I know that a lot of folks use EasyLanguage / Trade Station but I really don't want to try and learn yet another language.
Money is tight at the moment as the model which I was relying upon for my income has just blown up leaving me watching the pennies. (This was a very slow burn - non automated model and I have resolved that I need to get into shorter time frames and more automated to remove the emotion). In the past I have had bad experiences with PATS, and TT - although I would be willing to give them a second go - but they are really expensive.
I would appreciate any thoughts as to which software package / approach might be good? And any other ones that I should look at.
I have currently developed a custom indicator in eSignal which appears to be working well. I want to try and test it out on a few different instruments and wondered what would be the best software. Ideally I would want to use the same software to test the historical P&L, optimise it, and put it into production.
There is a strategy functionality in eSignal but it appears to be not based on tick data so I am very wary that it would not give accurate results. When I tried to build a strategy in it a few years ago I found that there was poor documentation, but I got something which might have been working but it was hard to verify. I am not great in JavaScript but am getting a bit better. I also a bit wary of using signal as a trading package - as it has its origins a charting packing but am not sure if this is justified (perhaps this is is due to bad experiences with CQG integrated client when it came out).
I also have an open e cry account (which I don't use much but have a free demo as I have left a little bit in their account) which has real tick data. When I was mucking around a few years ago I have worked out how to code some similar indicators in C#. These would be easy to adapt and in theory to create a strategy to test. But it was really frustrating to code in a language which I am not familiar with. OEC can also autotrade through into Excel through DDE which I understand is old technology and best avoided if I am starting from scratch. (I am most comfortable in VBA).
Does anybody know if OEC can see the code that I would be using for my autotrading / indicator (As they are the broker they can see who are the successful clients - or am I being paranoid?) Other negatives are that I am limited to just one broker.
I also have tried Ninja Trader in the past but this couldn't handle exchange based spreads and I couldn't work out how to export the data to check that my code was correct using Excel.
For my trading at the moment, I use Easyscreen which can autotrade through RTD Excel but it is expensive (I was using it for the autospreader which I probably won't need if this model works as I can get away from synthetic spreads). Although this is robust and properly supported - it won't help me with the testing and optimisation and would only really work for putting the model into production. If I went down this route I would have to buy the tick data and analyse it in Excel / VBA.
I know that a lot of folks use EasyLanguage / Trade Station but I really don't want to try and learn yet another language.
Money is tight at the moment as the model which I was relying upon for my income has just blown up leaving me watching the pennies. (This was a very slow burn - non automated model and I have resolved that I need to get into shorter time frames and more automated to remove the emotion). In the past I have had bad experiences with PATS, and TT - although I would be willing to give them a second go - but they are really expensive.
I would appreciate any thoughts as to which software package / approach might be good? And any other ones that I should look at.
Last edited: