was orders of magnitude harder than i'd ever thought... aside from the decision rules (ie the system, the algorithm) themselves that took me a full year of hardcore research, the coding, testing and backtesting part (multi-year tick data) was like climbing mountains without map nor compass in an unknown land... fortunately, i'd done that before! Launched Aug 8 noon JST (Japan Standard Time) and clippin' away... yeahhh!!!
If we are not talking 100s per day we are not talking high frequency by today's standards. Good luck with things 2cents! As I am sure you have already realized from your testing, signals are easy to generate, it's avoiding the chop and not giving it all back in commissions and slippage. Keep us all posted on your success. Auto
thanks Auto - i am in the low 100s at this point, and will remain there for a while till all the right parameters align to justify higher frequencies for this system
hmmmm, i always thought high frequency was 100's of trades a minute--- i guess it depends on who you talk to. best, surf
only one master model used to generate what one could call "sub-models" according to each instrument's "signature" matrix... trading 3 currency pairs right now but can increase the count as desired with a couple of mouseclicks