I'm looking to start a team of developers, quants, and financial experts, to setup and manage an auto-trading-money-making-machine

ryguy7272 ryanshuell at gmail.com
Tue Oct 14 14:16:43 CEST 2014


I'm looking to start a team of developers, quants, and financial experts, to setup and manage an auto-trading-money-making-machine

#1) Setup a VM; must be a Windows machine (maybe Azure)
#2) Load & configure the software (Matlab, Python, R, Excel, and SQL Server)
#3) Develop and test code (Matlab or Python....most likely)
#4) Hook into trading tool for order execution (system to be determined; depends on API)
#5) Setup a corporate business structure (this is virtually done, we just need to flip a switch)
#5) Shop around for clients (if we have a real money-making machine, this should be super-easy)


Just so you know, I've done this kind of thing before, using Excel and VBA, as well as an execution program called Sterling Trader. It was quite profitable, and typically had less than 1 down day in a 22-trading-day month. The system was profitable about 95% of the time. However, I don't want to use Excel for this project; I want this to be a real system. I think we can use Matlab, or Python. If this is project turns out to be very profitable, I know we can raise capital very quickly and very easily. At the beginning, I believe it will take a fair amount of work, but if we put in the effort, we can have a system that constantly monitors the equity markets during trading hours, finds the best profit opportunities, according to the trading strategies that we deploy, and once it is setup and running, we really won't have to do a whole lot. Ideally, once you turn it on, the whole process will require almost no intervention. I know this can be done; many banks have groups that do exactly what I'm proposing here. The top hedge funds do this too.

In conclusion, I think the trading strategies that we choose to employ should be easy to setup and test (a well-defined Google search will reveal countless trading strategies). I think getting the data should be pretty easy as well (we can load all kinds of indices into the tool). I think the hard part will be developing the automation processes; the tool will have to follow MANY rules and run all by itself. Finally, as neither Matlab nor Python are really execution tools, we'll have to connect to some type of system that allows us to send trade orders. This may, or may not, present somewhat of a challenge.


As an alternative, if we decide we don't want to manage money for other people, we could setup the business as a subscription service, and charge users, let's say $25k/year or whatever, and let people run simulations in our hosted environment, and they can manage money themselves...we simply provide all kinds of analytic tools for them to do what they want to do.



Hopefully everyone is close to New York City or Washington DC, or somewhere close, like Boston or Philadelphia.



More information about the Python-list mailing list