Some codes to do computational fincance are shown in this folder.
For AvellanedaLee 2009 , you need a matrix ij, called A, where j is different stocks to be considered (hardwired at 471) and i is the number of observations (hardwired)
I'll be posting updates with the threshold, signals and backtesting. Signal optimization is on the way. Remove hardwires.
- Anton Fadic
You are free to use any of these codes if you acknowledge the source.