In putting together my own backtest system, one issue I have ran into is the following.
Background. I use the alpari 1m tick data and occassionally there are gaps in data where no trades went off. Right now I fill forward "reasonable" gaps with the most recent 1min data when I load the data (setting open=high=low=close w/ vol=0).
Question. Should I give my backtest system the opportunity to trade those "filled" ticks? Can I assume any level of liquidity at that b/a even though no trades went off historcally during that time period (vol=0)?
Background. I use the alpari 1m tick data and occassionally there are gaps in data where no trades went off. Right now I fill forward "reasonable" gaps with the most recent 1min data when I load the data (setting open=high=low=close w/ vol=0).
Question. Should I give my backtest system the opportunity to trade those "filled" ticks? Can I assume any level of liquidity at that b/a even though no trades went off historcally during that time period (vol=0)?