Automatic Statistical Arbitrage (Matlab / IG Markets)

In November 2015 I though about bringing some results related to stochastic network theory into financial information processing. The idea was to develop an information filter which can be used as a very short term trend predictor in highly-volatile contexts. As I planned to test my results against empirical data, I needed market data and a reasonable filtration of "financial information" signals.

Good financial data (i.e. reliable and dense) is typically a luxury. Because I needed it at tick-by-tick level, I eventually resorted on writing a MATLAB listener to fetch free real-time data from a popular CFDs Market Maker. Apparently, this is the first automatic MATLAB API for IG Markets available online. So, please take it with a grain of piety.

As I realized I could easily saturate the stream allowance with no major performance loss (i.e. the script can open up to 40 contemporaneous push lines), I decided to complement it with an operating component which may handle up to 100 orders per minute.