Is this test reliable? - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
The image below is a backtest of one advisor in the mql market that I bought recently:
Drawdown is under 30%, the question is could it be manipulated?
You bought it but you ask the opinion to others, after the fact? Oh man!
It's nice to believe that water can be turned into wine. Someone did it but he his the same who said that you need to forget about money :-D
Seriously, I want to give my personal opinion based on my experience: never go live if you don't know which is the concept behind the strategy and the reason why it should work over the time. Back tests says actually nothing, do not believe them (apart from the fact that a back test can be rigged with sundry filters to hide streak of losses and curve fitting). I give you an easy example: ADX show that the trend is sustained right now (but it is actually marking the past) so I assume that I can entry expecting that the market will keep that same type of behavior for the next X period of time. Well, who tells me it will do it? I can only assume that there is a probability that it will persist in that way but...who know? This is the tragedy of traders totally dependent on algorithms; yes you can use robots sure, but only to assist your job and by the way the hard job is always discretional and based on your own judgment. This is how succesful traders/investors thinks: if I don't (or can't ) realize the principles and the why a thing should work so it is not convenient for me to invest money in that idea.
Backtesting is OK, just follow the rules below:
1. Backtest with a large spread (30 points or at least 1.5-2 times the usual spread of the desired pair) for the whole period of 21 years allowed by the platform.
You will usually get large drawdowns for October 2008 (Lehman Brothers), April 2015 (I don't know the reason) and March 2020 (COVID-19).
2. Try various settings and timeframes, nobody cares to produce fake results for 21 years and multiple timeframes. Don't buy the 'market has changed' claim. Not that much !!!
3. Double the maximal drowdown you get, 30% in your case, so calculates your lot considering a 60% drawdown.
4. Be sure to leave enough room in your account for margin + news related spikes + large spread at the start of the trading day + any adverse movement of 50-70 pips.
5. Enjoy !
Backtesting is OK, just follow the rules below:
1. Backtest with a large spread (30 points or at least 1.5-2 times the usual spread of the desired pair) for the whole period of 21 years allowed by the platform.
You will usually get large drawdowns for October 2008 (Lehman Brothers), April 2015 (I don't know the reason) and March 2020 (COVID-19).
2. Try various settings and timeframes, nobody cares to produce fake results for 21 years and multiple timeframes. Don't buy the 'market has changed' claim. Not that much !!!
3. Double the maximal drowdown you get, 30% in your case, so calculates your lot considering a 60% drawdown.
4. Be sure to leave enough room in your account for margin + news related spikes + large spread at the start of the trading day + any adverse movement of 50-70 pips.
5. Enjoy !
Thank you for your reply.
I hope you will come and tell us of the results.
This advisor was a complete disaster and was removed from the market.
This advisor was a complete disaster and was removed from the market.
This is what I m afraid for all these 5 stars EAs.. Some people say that backtesting in tester is not reliable but I cannot understand why. Theoritically we are testing an EA in a specific period of 6 month or 1 years in various symbols at the tester with "every tick.." method. If it works in tester, isn't safe to say that this EA is good enough ???? I have found one, tested in 3 different periods of 6 months each one, and gave good results. Why some say that tester is not reliable ? I m trying to understand.