An Improved Collective2

G

gmst

There are a lot of problems with Collective2. I would like to setup a clone website of collective2, which will have better features, lower fees, proven systems which are allowed entry into the website only after some due-diligence etc. This project will also help me increase my understanding of software engineering issues. :)

To start, a basic version of the website should have following features:
1) One webpage for each system
2) An order entry form and a http GET based order entry API
3) Auto-trading facility with 1-2 brokers as a start
4) A back-end database to store all systems trade history
5) Code to compute trading stats, and code to display performance graph
6) Code to query systems from the database (like Grid in collective2)
7) Code to show current and expired subscribers to a trading system and the money that will be transferred to the system owners.
8) Website will have to undergo due-diligence with CFTC and I assume the process might take 2 months???

I will be interested to know if someone can give me an idea of how many software Man-hours and weeks/months will be needed to achieve something working like above? My options are:

1) Hire one computer science engineer (fresh out of college) for this project. The guy will work 70-80 hours per week on this project.
2) Hire 2-3 computer science engineer(college students) as interns. Each will work 20 hours per week on this project.
 
Apart from the 8 things I mentioned above, feel free to add what are the other MUST HAVE features in a basic working version 1.0 of the website that I might have missed.
 
Version 1.0 will only trade futures and only with market, limit and stop orders. No OCO permitted in version 1.0. Basically, my question is how fast can I get something like this going?
 
Quote from gmst:

There are a lot of problems with Collective2. I would like to setup a clone website of collective2, which will have better features, lower fees, proven systems which are allowed entry into the website only after some due-diligence etc. This project will also help me increase my understanding of software engineering issues. :)

To start, a basic version of the website should have following features:
1) One webpage for each system
2) An order entry form and a http GET based order entry API
3) Auto-trading facility with 1-2 brokers as a start
4) A back-end database to store all systems trade history
5) Code to compute trading stats, and code to display performance graph
6) Code to query systems from the database (like Grid in collective2)
7) Code to show current and expired subscribers to a trading system and the money that will be transferred to the system owners.
8) Website will have to undergo due-diligence with CFTC and I assume the process might take 2 months???

I will be interested to know if someone can give me an idea of how many software Man-hours and weeks/months will be needed to achieve something working like above? My options are:

1) Hire one computer science engineer (fresh out of college) for this project. The guy will work 70-80 hours per week on this project.
2) Hire 2-3 computer science engineer(college students) as interns. Each will work 20 hours per week on this project.
More...

There are websites where you can hire feeelancers to do a specific project. I believe it's cheaper.
 
I also suggest that in version 1.0 no margin should be used as it's easier to implement.In other words trading stocks only.
 
The problem isn't setting up a website. The biggest issue is to find decent systems what people want to subscribe to...

Go to C2 and tell me how many systems do you see there what you would throw money at????
 
Quote from Pekelo:

The problem isn't setting up a website. The biggest issue is to find decent systems what people want to subscribe to...

Go to C2 and tell me how many systems do you see there what you would throw money at????
More...

I wholeheartedly agree.

In addition, I also believe there are a whole lot of issues with the C2 website and the way it is managed as of now:

1) C2 website regularly stalls during fast markets. They need to add servers but they are penny pinchers.
2) C2 allows developers to change names and in effect start a new life on website - thus it is not possible to know how many times their systems blew up before
3) C2 customer support is almost non-existent. You have to send multiple mails that get regularly ignored. You can get answers only if you put forward issues in the open forum.
4) Even after being in the business for 8+ years, so many damn things don't work on that website that it is simply sad. e.g. no margining for spreads.
5) Matthew - the guy who runs the C2 website appears more interested in funny jokes, writing novels (no disrespect intended, but it is as I see it), rather than developing and putting in more functionality to the website.
6) There is practically no oversight on quality of systems that are allowed on C2. Very recently there was a system called "Safe Capital" which after 3 consecutive months of 50%+ performance, blew up----losing 100% in the 4th month. Many such examples exist over the years on C2, and Matthew should have done something about better due-diligence, improved risk management etc. Alas, he choose to ignore.

7) Finally, the whole damn business model of C2 is take 98$ every 6 months (now it is 120$) from vendors, take 30% of subscription fees and allow new vendors with shitty systems to burn retail traders. The whole thing sucks and it is high time an alternative is provided to the investing community.

I have decided that I would be the person who will bring that alternative before the investing community. How hard the journey will be - we will know in due course. But I do intend to use this thread to document it.
 
Quote from gkishot:

I also suggest that in version 1.0 no margin should be used as it's easier to implement.In other words trading stocks only.
More...

Thanks for suggestion. Noted.
 
the way to fix some of the problems you mention would be to make it that all systems have to be in real accounts, not demos.
But how to make that run on one website might be problematic
 
You must log in or register to reply here.
Back
Top