Backtesting with tick data vs. live trading results - page 2

New comment
 

Forum on trading, automated trading systems and testing trading strategies

Backtesting with tick data vs. live trading results

Robert Jagger, 2020.09.25 15:41


Don't put too much trust in the term 'real ticks', since it is the coding of the EA and not the settings or data in the strategy tester that matters most. 

This video covers this topic well. https://www.youtube.com/watch?v=dGFphm2bFbs

Nice points in the video ,but 7 minutes just to say its better if you trade signals on new bar formations (which is what 95% of systems developed here can do already).

I have a lot of objections for the Open prices backtest for this video ,but i keep forgetting to examine how the test behaves and which prices are processed ,so i cant argue yet.

The theory for his ideas is well grounded though .

Edit :

tested open prices only on MT5 :

Test1 : How many ticks does open prices only send per bar => 1 with open price 
Tested on M1 ,got 1440 ticks exactly all 1 minute apart

2020.09.25 18:52:31.488 Core 1  2020.09.24 23:57:00   Tick [1438]Ask[1.16672]Bid[1.16671]Time:2020.09.24 23:57
2020.09.25 18:52:31.488 Core 1  2020.09.24 23:58:00   Tick [1439]Ask[1.16670]Bid[1.16669]Time:2020.09.24 23:58
2020.09.25 18:52:31.488 Core 1  2020.09.24 23:59:00   Tick [1440]Ask[1.16681]Bid[1.16680]Time:2020.09.24 23:59

Test2 : If we have a buy order and the Low of a candle punches below the orders stop loss is that processed in open prices only mode ? => it is 

If its a system SL/TP its processed ,if you do virtual stops you have to calculate that on your own

So in theory whether you test with simulated ticks or open prices on M1 ,is quite similar ,and as he said ,faster -since you are using M1 anyway.

Now if your system only has Hard SL and TP and it does not have :

  • trailing
  • break even
  • equity stops
  • equity trailing
  • multipair/multitf capacity

and you can adjust your ea to operate like in the open prices only test ,then theres no objections indeed .

(since the systemic SL fired in the test im assuming a possible margin call would fire as well)

 
I know hundreds of reasons why live trading results might be much worse than backtested ones. This is the usual case. But since I am observing the opposite, I run out of explanations. So far no help here unfortunately.
12
New comment