As I already said in my last article: I’m interested!
And for a trader that means: Let’s test it
But how todo it?
For a software developper it’s easy
As input the best source seems to be an ESignal Tickfile, which covers the last 10 trading days and contains every tick in an Ascii-file.
And to setup the Backtest-Tool my FuturesTrader software already has most of the necessary routines already build in. So it was just a few hours work to build the first implementation of a System, which takes a trade always with the trend and reverses positions, if the trend changes.
For the time being I use 1 tick/trade for commission and slippage. In my real trading I usually have no slippage as I trade with Limit and Stoplimit orders only. So 1 tick/trade should be sufficient, but a later implementation might use Bid/Ask information contained in the Tickfiles for Entry and Exits.
While on ES in 10 days the system lost 2 ticks overall on 16 trades with no trailed stop or at least a stop moved to Breakeven after a certain amount of profit, on the Euro the results weren’t as good:
44 ticks lost on 22 trades.
But there is a lot of room for improvement on the system and as I said, I’m interested and curious, if this is the way I can find a profitable autotrading system.
Future implementations will consider a Breakeven +1 Stop, a trailed Stop, once the position is in the profit, taking targets and most important: Statistics, so it becomes easier to compare different sets of parameters. Sure all this is finetuning and may fit only the data I have, but you have to start somewhere. And I might expand the range of acceptable input files in the future, so the program is able to read minute barfiles, which can go back over a very long time. But for the time being I’m happy to have the ESignal tickfiles, which are all that’s needed to backtest current market conditions.