I have a question: I found a very usefull indicator. But this indicator is multitimeframe and repaints. Its name is RSI 3TF. How can I backtest such a indicator? I have the feeling, that in backtest mode (Strategytester MT4) other arrows appears, as in real life. Is this possible? For example: The indicator repaints, and because of this I saw in real life that arrow appears and later disappears. But in strategytester this “false arrow” doesnt appear. Is this possible? I thought that the strategytester (backtesting) mode shows the same results or arrows as in real life only in the past. Or does the MTF indicator shows in the “future” when backtesting?
My personal opinion is that back testing an indicator that repaints is not possible as it will be inaccurate. And inaccurate data is pointless. You can’t even go back through the numbers manually to check for errors.
But why is this so? I dont understand, because normally the strategy tester should only take the past data isn´t it? That means the data which the real time chart also used. That means, that the indicator, doesnt matter in backtest or in real chart, should show the same repainting arrows or? I dont understand. Why the indicator should show other arrows in backtesting?
The problem here is that only the creator/programmer of the EA knows exactly how the EA should or will behave, you as the end user do not know the different mathematical parameters the EA has to go through to get to the end result. Hence when doing back testing you can’t be sure when you see a trade if that trade is taking place before a re-paint or after, I would assume the latter, but since I didn’t create the EA I cannot be sure.
I understand your frustration, I had a similar problem with an indicator about a year and half ago and got rid of it as it wasn’t reliable, it only showed good trades in hindsight.