Finance has always fascinated me. It is ripe with mathematics, very hands-on, it has a global marketplace, the assets are valued all the time. Other interesting aspects are big data, complex relations and the possibility for endless challenges as the market evolves. It is a field perfect for trying out machine learning technology, and who knows maybe hit jackpot if the findings are profitable. But that is not an initial goal.
The goal for me is to set up a platform that allows me to build different trading algorithms and evaluate them.
Initially(this article), I want to
- Find a python library to support building and backtesting algorithms
- Setup an evaluation method to evaluate the performance of a strategy
- Construct a simple trading algorithm to showcase the evaluation
- Run the system on my own laptop on demand
Further down the line I want to
- Have a system that can generate trading signals in different markets
- Run the system on AWS and update automatically
- Have a web frontend which shows the performance of the algorithm(s) and the signals
- Have the algorithms connected to a real account to do automatic trading – far into the future
Of course, this is not an exhaustive list, and many more aspects of it will, without a doubt pop up. So keep reading.