Profit Generator EA - page 32

New comment
 
amckeen:
I am in the red (negative) so far this week, testing seven pairs on two settings.

There is, however, a bright side: if I backtest over the same time period (since market open last night), I get almost identical results. Much of the time even the entry point and time match exactly between the back and forward tests.

To me, this implies that backtesting is pretty accurate, and that we're just in a slump. If we hang in there, I expect we'll soon see the results that deeper backtesting indicated.

Anyone in the black for today?

Its all red here also, could not understand why all red, only 1 or 2 turned green, is there a way to avoid days like today?

I wish if we could back test on tick data to be more accurate...

 
amckeen:
I am in the red (negative) so far this week, testing seven pairs on two settings.

There is, however, a bright side: if I backtest over the same time period (since market open last night), I get almost identical results. Much of the time even the entry point and time match exactly between the back and forward tests.

To me, this implies that backtesting is pretty accurate, and that we're just in a slump. If we hang in there, I expect we'll soon see the results that deeper backtesting indicated.

Anyone in the black for today?

That is good to know. Why don't we all together focus our efforts on backtesting both short and long term and see what we come up with. We all need to backtest different settings and try out different settings. That is what we all need to work together as a group.

You are right, I don't mind losses as that is part of trading. However, currency pairs to change they "way" they trade. Let's focus on finding good short term settings via backtests on these. Please, let's focus on backtesting like crazy and see if we can come up with something.

 
laserjet:
Its all red here also, could not understand why all red, only 1 or 2 turned green, is there a way to avoid days like today? I wish if we could back test on tick data to be more accurate...

I have noticed that trending days don't do so well with this EA. However, as a general rule, there are not too many trending days so overall the EA does fairly well. I do think there might be a way to limit the downside but we need to all do some serious backtesting to find those settings.

 

If we are to be successful traders, we will have to accept losers. There is no way around it. Too many filters will weed out the losers but more often than not, it will weed out more winners and make the system less robust. If one day of losing and that too in a demo system troubles you, then step back and take a hard look at your wishes to be a trader. I don't mean to sound harsh, but please think about it. We are trying to find winning combinations, not the holy grail. There will be losers and strongly trending markets will cause "dip buyers" which this system is, to lose money. However, as holyguy says, the market trends only a fraction of the time, so overall the system should make money. As for trending times, one should also have some breakout/trend following systems to compliment dipbuyers. It is the principle of diversification, which is so essential for a smoother equity curve.

Good luck everyone.

 
Maji:
If we are to be successful traders, we will have to accept losers. There is no way around it. Too many filters will weed out the losers but more often than not, it will weed out more winners and make the system less robust. If one day of losing and that too in a demo system troubles you, then step back and take a hard look at your wishes to be a trader. I don't mean to sound harsh, but please think about it. We are trying to find winning combinations, not the holy grail. There will be losers and strongly trending markets will cause "dip buyers" which this system is, to lose money. However, as holyguy says, the market trends only a fraction of the time, so overall the system should make money. As for trending times, one should also have some breakout/trend following systems to compliment dipbuyers. It is the principle of diversification, which is so essential for a smoother equity curve. Good luck everyone.

I totally agree with you Maji, backtest also show lot of loses, loses are part of trading. I know there is no holy grail, if every one start making money who is going to lose. Anyhow what I asked how to backtest on tick data to be accurate....

 

curve fitting?

Hi, just a thought... I see settings in the posts like LongBar 16, 28, etc.

isn't that pretty much like curve fitting? I mean... LongBar 10, 15, 20, etc. makes sense but is a 16 really different than a 15? I mean it might produce better results in backtests but in forward tests I don't see how we could expect a 1 point difference to mean that a 16 is consistently better than a 15 or vice versa.

disclosure: I haven't been following the thread completely so I don't know exactly what I'm talking about but just thought it would make a valid point.

 

PG, which time period bar chart are you suggesting to use on?

Hi there,

Which time period do you suggest using on?

thanks

E

 

Forward Test Profit Generator 3[1].1

I setup a forward test on FXDD for the new Profit Generator 3[1].1 with all pairs H1 canned settings.

http://www.thehobbystop.com/profitgenerator3/statement.htm

I'll put a link on the website.

 

I haven't optimized it, but I think OpenBarPercent is going to be as important as LongBar.

1...252627282930313233343536373839...85
New comment