Automate IS and OOS Backtesting

[Ironman]

I would like to automate the following backtesting routine:

-> Run an optimization on a specific time period, the in sample data
-> Pick the the five strategies with the highest Net Profit from that optimization
-> Run those five strategies on a different time period, the out of sample data

Can this process be automated with AmiBroker and the Batch Window or somehow else?

Thank you

[Tomasz]

You find best strategy using built-in optimizer:

strategy = Param("Strategy", 1, 1, 5, 1 );

switch( strategy )
{
  case 1:
   // strategy 1
    Buy =
    Sell =
   break;

  case 2:
   // strategy 2
    Buy =
    Sell =
   break;

  case 3:
   // strategy 3
    Buy =
    Sell =
   break;


  case 4:
   // strategy 4
    Buy =
    Sell =
   break;

  case 5:
   // strategy 5
    Buy =
    Sell =
   break;
}

And you can run built-in walk forward to switch between best strategies in OOS periods.
If you wanted to sum 5 "best" strategies, you could run just 5 WF sequences successively eliminating those top in subsequent runs.

[Ironman]

Thank you for your help. I think I described what I wanted to do not good enough.

With your process I have to put the strategy code into a code. That is not what I want.

I have allready a strategy with a lot of functions and parameters to optimize. The result of the optimization (non-exhaustive opt.) are hundrets of different variations with certain parameters.
I would like to pick the top five or top ten. (Net Profit).
And then run the stratety with the different parameters on the out of sample data.

I am doing this proecess right now, manually. But I want to automate it somehow. I thought on the batch window.

Thank you very much

[Pinecone]

This thread might be useful for you.