How do you divide your time between backtesting historical charts and forward testing in current market conditions? Can a strategy that look
How do you divide your time between backtesting historical charts and forward testing in current market conditions? Can a strategy that looks excellent on historical data still fail when traded live?