A question for those people who have written their own back testing frameworks and automate though IB, do you use the IB historical data to backtest or do you source tick data from somewhere else? If you source other data do you not run the risk of optimizing systems which may exploit data quirks which may not exist in the live IB data?
