Goblin "BiPolar" Edition - page 47
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
Simulation of using progressions
Hi Rico,
Interesting idea you have.
Since I am not completely sure that I understand it clearly in every details of it, would you mind posting an example with figures that would demonstrate your theory?Here are some backtest results for year 2001:
buy winners 544, buy losers 25, sell winners 566, sell losers 25
total winners 1110 (95%), total losers 50 (5%)
longest run of successive winners 87, shortest run 2, average run 23.1
The first graph shows standard fibonacci lot sizing for orders, every trade starts with the same conditions.
The second graph shows the same trades, but with an inverse progressive lot size after a loser trade (start with factor 7, go down after each winner until factor 1 is reached again).
The third graph show the same trades, but with a progressive lot sizing, meaning that after a winner, the factor increases from 1 until finally 5, a loser restarts the progression.
As you can see the differences are interesting. Basically I am looking for a way of determining the best lot sizing method that exploits the statistical regularities of winners vs. losers.
update last 2 weeks
update last 2 weeks.
And here is back test with some settings changes to the risk and equity target, i'll be running next week.
DaveInteresting results, which settings are you using?
Thanks, Rico
Have been testing "target equity to close" and my conclusion is to start off with value at 50% starting equity say starting equity=1000 then target equity should be 1500, then change to 2000, then to 2500 in 500 increments and so on as trend continues.
Here are some backtest results for year 2001:
buy winners 544, buy losers 25, sell winners 566, sell losers 25
total winners 1110 (95%), total losers 50 (5%)
longest run of successive winners 87, shortest run 2, average run 23.1
The first graph shows standard fibonacci lot sizing for orders, every trade starts with the same conditions.
The second graph shows the same trades, but with an inverse progressive lot size after a loser trade (start with factor 7, go down after each winner until factor 1 is reached again).
The third graph show the same trades, but with a progressive lot sizing, meaning that after a winner, the factor increases from 1 until finally 5, a loser restarts the progression.
As you can see the differences are interesting. Basically I am looking for a way of determining the best lot sizing method that exploits the statistical regularities of winners vs. losers.When will you all realize MT4 backtester DOES NOT WORK. Publishing your backtests is about as uselss as using a fortune cookie to predict market movement.
-secXces
When will you all realize MT4 backtester DOES NOT WORK. Publishing your backtests is about as uselss as using a fortune cookie to predict market movement. -secXces
so? what do you suggest, blindly buy more computer to run forward test? i roughly check over TSD forum, found about 1400 version of EAs around. Say, you run 20EAs on 20 platform with 1 settings on 1 computer, you'll need about 700 unit of computer just to figure out the default setting is work or not for the 1st few months. stop posting negative!
regards,
David
Here are some backtest results for year 2001:
buy winners 544, buy losers 25, sell winners 566, sell losers 25
total winners 1110 (95%), total losers 50 (5%)
longest run of successive winners 87, shortest run 2, average run 23.1
The first graph shows standard fibonacci lot sizing for orders, every trade starts with the same conditions.
The second graph shows the same trades, but with an inverse progressive lot size after a loser trade (start with factor 7, go down after each winner until factor 1 is reached again).
The third graph show the same trades, but with a progressive lot sizing, meaning that after a winner, the factor increases from 1 until finally 5, a loser restarts the progression.
As you can see the differences are interesting. Basically I am looking for a way of determining the best lot sizing method that exploits the statistical regularities of winners vs. losers.Alassio,
Being Bipolar Edition under discussion here, during trading occasionally both sell and buy orders exist at the same (period) of time. How does your version handle this situation? Does the lots sequence of the "losing" side remain the original one (while these orders are losing position)?
Regards,
Chrisstoff
Alassio,
Being Bipolar Edition under discussion here, during trading occasionally both sell and buy orders exist at the same (period) of time. How does your version handle this situation? Does the lots sequence of the "losing" side remain the original one (while these orders are losing position)?
Regards,
Chrisstoffthe whole concept behind bipoler is to have it trade both directions at any given time. all other martingale style ea's place orders in one direction untill max trades are met. Then waits for the last trade to get hit. Bipoler trades the opposite direction while waiting for the other orders to close out.
positions and management are the same regardless of open orders.
and as for the guy pissin on back testing.......back test are mearly to give a person an "idea" of how an ea will run, to see that the functions with in the ea are working, and is a valuable aid in the developement.
But notice that I have posted more forward test than back test and all are very good results on a weekly basis.
Update on live account: account is open and awaiting funding. Soon as my check comes.
first deposit 3500 and I got another 2500 going in shortly after, as the last month trading has been very good. I only make withdrawls on a monthly basis.
Dave
the whole concept behind bipoler is to have it trade both directions at any given time. all other martingale style ea's place orders in one direction untill max trades are met. Then waits for the last trade to get hit. Bipoler trades the opposite direction while waiting for the other orders to close out.
positions and management are the same regardless of open orders.
and as for the guy pissin on back testing.......back test are mearly to give a person an "idea" of how an ea will run, to see that the functions with in the ea are working, and is a valuable aid in the developement.
But notice that I have posted more forward test than back test and all are very good results on a weekly basis.
Update on live account: account is open and awaiting funding. Soon as my check comes.
first deposit 3500 and I got another 2500 going in shortly after, as the last month trading has been very good. I only make withdrawls on a monthly basis.
DaveI haven't had to much luck trading with this EA. What version and settings are working for you? I would like to forward test this EA again.
Thanks
so? what do you suggest, blindly buy more computer to run forward test? i roughly check over TSD forum, found about 1400 version of EAs around. Say, you run 20EAs on 20 platform with 1 settings on 1 computer, you'll need about 700 unit of computer just to figure out the default setting is work or not for the 1st few months. stop posting negative!
regards,
DavidFirst Off, I am part of a community that runs around 100 tests a week on multiple different ea's. Your daddys computer is probably capable of running around 5 versons of metatrader all by itself. And if you knew anything, you could do spread analysis. Oh too big of a word, rephrase, you take the absolute probable low of a setting and the absolute probable high, and then you take the spread, adjusting the testing to the closer side of your positve outlook. Again, your backtesting is useless, the only thing it is used for is to test the EA to make sure it doesnt have any bugs in it. Run it live for a week, then backtest that week, check your results. Second, by defualt setting I assume you mean adjusting it too that perfect magical number right? You sound like one of those people, well, this well help you out. Go to google and search for curve fitting. Learn it. Then you can understand how funny that smart elec comment is. Third, im by far being negative. Im simply pointing out some things that need to be said. Negative would imply something derogatory or offensive to someone. If you felt that way, then you need to do a check up from the neck up bub. My apologies, you know, what I said I wasnt going to post on this thread anymore. This is my fault. I apologize. By all means, continue the backtesting.
- secXces
First Off, I am part of a community that runs around 100 tests a week on multiple different ea's. Your daddys computer is probably capable of running around 5 versons of metatrader all by itself. And if you knew anything, you could do spread analysis. Oh too big of a word, rephrase, you take the absolute probable low of a setting and the absolute probable high, and then you take the spread, adjusting the testing to the closer side of your positve outlook. Again, your backtesting is useless, the only thing it is used for is to test the EA to make sure it doesnt have any bugs in it. Run it live for a week, then backtest that week, check your results. Second, by defualt setting I assume you mean adjusting it too that perfect magical number right? You sound like one of those people, well, this well help you out. Go to google and search for curve fitting. Learn it. Then you can understand how funny that smart elec comment is. Third, im by far being negative. Im simply pointing out some things that need to be said. Negative would imply something derogatory or offensive to someone. If you felt that way, then you need to do a check up from the neck up bub. My apologies, you know, what I said I wasnt going to post on this thread anymore. This is my fault. I apologize. By all means, continue the backtesting. - secXces
It all depends on how the EA is programmed. If you always calculate it on previous bar and don't use small take profit/stop loss. The Back tester is ok.
Or it is useless. Did anyone test it with tick data?