Automated Trading Software

[Razorsharp1]

I have been trying to develop a trading system but have had a few problems with the programmers.

I have some very good strategies that i wish to automate and am looking for someone who may be interested in joining me. I have listed below some of the requirements:

Algorithmic trading platform with the following components:
- User Interface
- Portfolio Management System
- Order Management System
- Database design
- Historical data
- Testing framework for back-testing algorithms on historical data
- Reporting and charting
A partial overall system design and partial code exists for this project. The programmer will be required to spend time assessing the current design and implementation; to undertake a gap analysis which sets out the difference between the desired system and the current partially-implemented system; to provide an implementation plan including if possible an estimate of required effort; and to redesign and implement the missing parts of the system to create a working whole.
Experience/knowledge about the following will be needed:
• Essential: Experience in development in C++ and Javascript
• Essential: Database access, database design and optimisation for high performance
• Essential: understanding of Order Management System and Portfolio Management Systems
• Essential: ability to produce structured reports and charts both for the user interface and for printout
• Preferred (or able to learn quickly): Connecting to a broker via FIX/ broker API
• Preferred: Implementation of Order Management System and Portfolio Management Systems
• Preferred: Implementation/reporting of back testing engine



I am open to discuss various options for compensation.

Thanks

 

[BeginnerJoe]

I'll play if the money is good. I am curious though, what is your expectations of how long this would take ? Also, are your strategies known to be winners ? I would be less interested in excessively experimental/exploratory type of projects because non-development factors would then get in the way and become a hindrance.

 

[Razorsharp1]

I'll play if the money is good. I am curious though, what is your expectations of how long this would take ? Also, are your strategies known to be winners ? I would be less interested in experimental/exploratory type of projects because non-development factors would then get in the way and become a hindrance.

Thank u Joe. I have an idea of how long this should take. My expectations are around the 4 weeks mark, give or take. What are your expectations of good money ?

Strategies are winners, on auto their performance is only set to improve.

 

[BeginnerJoe]

4 weeks for a money making machine is insanely optimistic. It would probably take that long for some relatively good programmers just to get to grips with the partial system that you have. But, why 4 weeks ?

 

[Razorsharp1]

I don't think it would take too long for someone to get familiar with it. I have had two programmers, one from Morgan Stanley and one from Saxo Bank . The guy from Morgan Stanley went through it in two days but didnt work on it as he said he was too busy, but he did help the second guy a little. The second guy setup the database and started working on the rest but was way to slow in getting back and meeting deadlines. He couldn't get it working but managed to create a separtate platform that collected the data.

I am wary of programmers looking to take the project on as a wage paid job, they tend to drag it out as long as they can. I would prefer to give the project on a phase completion basis.
Which part of the description above do you think would be the most time consuming?

 

[BeginnerJoe]

I am wary of programmers looking to take the project on as a wage paid job, they tend to drag it out as long as they can.

This may or may not be true. It's equally possible for people to underestimate what it takes to build systems. If a money making machine can be built in 4 weeks or 4 months, everyone would be doing it. Without knowing what your strategies are, what data they need, what computations for order generation, what computation capacity is available, it's difficult to say how long it would take.

Which part of the description above do you think would be the most time consuming?

The FIX connectivity would probably take a whole team of Morgan Stanley guys several years to make working reliably. That is not to say there aren't more capable people out there .


Trading system building is non-trivial and very much an investment. If the trading strategy is sound, the investment will pay off with the right people. If there is no strong confidence in the strategy, building a system will be a looser (I like the phonetics). This is why I asked how good your strategies are to assess your staying power for the investment.

 

[tedgregory]

That's great having a good strategy, tested one.
As to a trading robot: what are the key concerns for you?

One man can't physically be good at all skills you've listed. We have a team that is certainly capable to do everything you need. Let's get into details.

 

[tanyach]

Just curious. What platform did you decide to use?

 

[Razorsharp1]

This may or may not be true. It's equally possible for people to underestimate what it takes to build systems. If a money making machine can be built in 4 weeks or 4 months, everyone would be doing it. Without knowing what your strategies are, what data they need, what computations for order generation, what computation capacity is available, it's difficult to say how long it would take.

Is not a case of 4 weeks to build a money making machine. The program is just a tool to execute what could have taken years to develop. Without the correct systems the program will not make a penny. The industry is full of people who can build such programs from scratch yet they are working for someone who has the systems.

The FIX connectivity would probably take a whole team of Morgan Stanley guys several years to make working reliably. That is not to say there aren't more capable people out there

Maybe but not for the level of program I am looking to have done at this point. There are lots of small traders running small automated systems that are quite stable. Just looking for something like that at this point.

Trading system building is non-trivial and very much an investment. If the trading strategy is sound, the investment will pay off with the right people. If there is no strong confidence in the strategy, building a system will be a looser (I like the phonetics). This is why I asked how good your strategies are to assess your staying power for the investment.

I believe the strategies are good and will make consistent money. The program will make it easy to do the maths and also identify many more opportunities. I have spent a long time developing these systems.

 

[Razorsharp1]

That's great having a good strategy, tested one.
As to a trading robot: what are the key concerns for you?

One man can't physically be good at all skills you've listed. We have a team that is certainly capable to do everything you need. Let's get into details.

Thanks.

Not sure what you mean by concerns.

I am hoping that one man can do it

 

[Razorsharp1]

Just curious. What platform did you decide to use?

Not using any platform, had planned to make our own custom one, but it never got completed. Have a partial platform which we wish to complete.

 

[BeginnerJoe]

One man can't physically be good at all skills you've listed.

Well, it depends on who the man is. Some men are quite talented. You need luck to find such a man. Perhaps you have not been lucky.

But you hit the nail on the head for the OP. He will need a whole team of ordinary men for what he wants to do. But I am beginning to suspect he's looking for something that will run on a laptop. A whole team of men wouldn't fit on such a computer.

 

[Razorsharp1]

I don' think you are quite clear on the requirements. There are many traders, many on here who run automated trading systems which they developed themselves. I am sure they didn't rely on a team of programmers, they just knew how to create a lot of the requirements I have listed.

Either one can do it or one cant. But without a shadow of a doubt it can be done by one man. I m sorry if you don't quite see that but you may be surprised.

 

[BeginnerJoe]

There are lots of small traders running small automated systems that are quite stable.

Doesn't seem plausible. Trading systems would be prohibitively expensive for small traders to own. If you know a small trader with a profitable system, perhaps you can ask him to share and split the costs. A trading system is not like Microsoft Word that you can take home, use, and set up for life. An interesting fact, even for programmers, did you know a program like Microsoft Word actually takes 100's if not 1000's separate individuals to make ?

I believe the strategies are good and will make consistent money.

To be worth building a system, you have to KNOW it is profitable. Just believing it would be insufficient probability for success.

I think you are looking more for something like an MT4 EA. This can be done easily by a lot of people. If you can demonstrate your strategy as profitable using real trades, I will be happy to build you an MT4 EA/Robot/"trading system" for free.

 

[Razorsharp1]

Whether it sounds plausible to you is irrelevant, the fact is that there ARE small traders running small programs and running automated systems.

Thank you for your offer of programming an "EA", but that is not what makes a trading system, it may be one component. Those that know about trading and trading systems understand clearly that it takes more than an EA.

I KNOW the system works. The person who takes on this task will understand why the likes of MT4 or Ninja trader are not suitable for what i wish to create.

This is not a job for an EA programmer. I can program basic EA's myself. This requires a little more. I think some of the skill-sets i have mentioned are out of your scope. But I am sure there will be someone who knows what I require and be able to do it.

 

[BeginnerJoe]

I think some of the skill-sets i have mentioned are out of your scope.

Your skillz set is a child's play. But you ain't getting 'em cheap from me . I look forward to hearing your success.

 

[Razorsharp1]

Your skillz set is a child's play. But you ain't getting 'em cheap from me . I look forward to hearing your success.

After all these posts telling me it wasn't possible, all of a sudden the skill set is child's play to you?

Who said anything about cheap?? I understand there is a cost but as i said before am aware of programmers looking for job security for a few months.

if you had the skill set I am sure you would have started the discussion on a different note. Thanks for your comments but you are more transparent then you my realize.

 

[BeginnerJoe]

After all these posts telling me it wasn't possible, all of a sudden the skill set is child's play to you?

I said it wasn't possible for your Morgan Stanley guy or a whole team of them.

The assessment was mutual. I was looking for a high probability set up - in terms of 'project' success. It takes money and other elements. So far I don't see the other elements.

I think what you need is some cheap coders to play with and for you to gain experience on what building system is all about. It's going to be a long road.

 

[Razorsharp1]

I think we are completely on different pages. When you refer to other elements I assume you are referring to the Trading System itself. If so then that is not of any concern to the programmer as they will not be passed on. The only thing the programmer has to worry about is completing what is needed and his compensation.

I am not looking for experience on what building systems is about, have been doing exactly that for the last four years. I understand the flow and the process that makes a good system. Just need a programmer to complete this stage.

The next stage will be more complex and cost more. I am fully aware of where we are at and what is needed and how to proceed.

Thank you for your feedback.

 

[malaguti]

did you know a program like Microsoft Word actually takes 100's if not 1000's separate individuals to make ?
.

Rubbish