Backtesting problem

New comment
 

Good day


Can anyone please assist.

I purchased a trading robot for Crash500. When I run a test on the bot it is showing me results of big profits but when I run on real account I take losses. Not sure if I am doing it correct as when I sit and look at the test I picked up that as it runs it collects profits but when I have a buy open and market drops it does not seem to collect that negative data. My balance will remain the same even after it took a loss due to market dropping

 
Dawid Zinserling: Can anyone please assist.I purchased a trading robot for Crash500. When I run a test on the bot it is showing me results of big profits but when I run on real account I take losses. Not sure if I am doing it correct as when I sit and look at the test I picked up that as it runs it collects profits but when I have a buy open and market drops it does not seem to collect that negative data. My balance will remain the same even after it took a loss due to market dropping
Run your back test with "real tick data". Due to the inherent slippage of the drops and spikes in Crash and Boom, any EA for them cannot be back-tested in any other way, except with real tick data.
 
You can consult with the seller about all the particularities concerning the robot.
 
Fernando Carreiro #:
Run your back test with "real tick data". Due to the inherent slippage of the drops and spikes in Crash and Boom, any EA for them cannot be back-tested in any other way, except with real tick data.
Thanks. Will try it tonight.
 
Dawid Zinserling:

Good day


Can anyone please assist.

I purchased a trading robot for Crash500. When I run a test on the bot it is showing me results of big profits but when I run on real account I take losses. Not sure if I am doing it correct as when I sit and look at the test I picked up that as it runs it collects profits but when I have a buy open and market drops it does not seem to collect that negative data. My balance will remain the same even after it took a loss due to market dropping

Hi, Dawid

I want to share my experience when purchased Trading Robot. the seller recommended for backtest first, it's not wrong, but what our historical data is valid?

I look at the backtest from the seller as a good result and also by myself, but when applied to real trading it does not represent the backtest.

my question is maybe my backtest data is not valid, so I bought dukascopy historical trading data to have good and accurate data.

OK, I backtest again but the result in the real trading is still not same as backtest.

and finally, my conclusion is I have tested robot trading using a demo account or small account balance for two or three months.

maybe some of the people said, it takes more time, yes it takes a long time but give accurate real trading condition.


please consider how much money you invest in a new trading robot that you just met

 
Dani Rukmana #:

Hi, Dawid

I want to share my experience when purchased Trading Robot. the seller recommended for backtest first, it's not wrong, but what our historical data is valid?

I look at the backtest from the seller as a good result and also by myself, but when applied to real trading it does not represent the backtest.

my question is maybe my backtest data is not valid, so I bought dukascopy historical trading data to have good and accurate data.

OK, I backtest again but the result in the real trading is still not same as backtest.

and finally, my conclusion is I have tested robot trading using a demo account or small account balance for two or three months.

maybe some of the people said, it takes more time, yes it takes a long time but give accurate real trading condition.


please consider how much money you invest in a new trading robot that you just met

Just my opinion guys. I started investing and day-trading in 1997, so I've learnt quite a bit.  Backtesting is a waste of time, forget it altogether. You get very pretty results from backtesting because it optimises the period of time you selected. Amazing profits that you can see growing over time in the charts of the results - but only if you can wind back the clock and use the method the backtester came up with to trade the period of time it analysed. That has no bearing on reality. It's just another way of selling software.  Remember the great California Gold Rush?  The people who made most of the money were those selling the Picks and Shovels.
New comment