EA ECN Forward Test 2 (Od jriechers)
Użytkownik usunął ten system.
EA ECN Forward Test 2 Omówić
Uczestnik z Sep 04, 2009
879 postów
Jan 08, 2010 at 14:40
(edytowane Jan 08, 2010 at 14:44)
Do u still intend to go live with this lot size setting ?
Jan 08, 2010 at 15:17
Im having a hard time getting this point across... its not about lot size, its about risk % per trade and right now this test is running at 10% per trade. The system automatically calculates lot size based on available margin, stop loss and risk %. With 10% risk per trade I end up with about a 30-40% drawdown in backtests and so far Im sitting at 33.88% drawdown which is acceptable for me. This level of risk and related drawdown results in 200k+ trades per pair over a 2 month period.
Now before everyone gets critical of the short testing period, I found that a PA/Breakout system success rate is directly related to the volitility and volume of the market. For instance, a year ago GBPJPY was trading 2-3 times as many pips per day as it is now, and that figure changes month to month also. Two months allows for a little stability while adjusting to the market conditions.
Every week I adjust one pair using 8 weeks of historical data and various levels of ATR. I will eventually figure out how best to put this into the system, but for now Im more comfortable doing it manually. Basically the most important adjustment related to market conditions is how ATR is related to the Stop Loss setting, the more the market naturally moves, the larger the Stop Loss needs to be to allow for natural swings, but still be close enough to close a bad entry. This also affects the TP setting, but to a lesser degree as all trades are automatically closed before Friday close (which artificially affects my win/loss %).
Now before everyone gets critical of the short testing period, I found that a PA/Breakout system success rate is directly related to the volitility and volume of the market. For instance, a year ago GBPJPY was trading 2-3 times as many pips per day as it is now, and that figure changes month to month also. Two months allows for a little stability while adjusting to the market conditions.
Every week I adjust one pair using 8 weeks of historical data and various levels of ATR. I will eventually figure out how best to put this into the system, but for now Im more comfortable doing it manually. Basically the most important adjustment related to market conditions is how ATR is related to the Stop Loss setting, the more the market naturally moves, the larger the Stop Loss needs to be to allow for natural swings, but still be close enough to close a bad entry. This also affects the TP setting, but to a lesser degree as all trades are automatically closed before Friday close (which artificially affects my win/loss %).
It is what it is, until its not, and then its something else.
Uczestnik z Sep 04, 2009
879 postów
Jan 08, 2010 at 16:13
Wouldn't u get rich quick enough😉 even with reduced risk per trade to lets say 2%? When running on live account u would sleep waayyyy better i think...
Uczestnik z Sep 04, 2009
879 postów
Jan 19, 2010 at 15:43
doesn't look good... whats up?
Jan 19, 2010 at 20:14
Yeah i know... I just finished re-optimizing last night and added an additional pair (EURUSD) that is more set up for scalping.
This is what forward testing is for :) to find the weaknesses and adjust accordingly. Overall, I attribute the recent losses to the market shifting gears after an extended holiday period. We shall see.
I have another account up trading 9 total pairs/systems on a live 1000 account... oh joy.
This is what forward testing is for :) to find the weaknesses and adjust accordingly. Overall, I attribute the recent losses to the market shifting gears after an extended holiday period. We shall see.
I have another account up trading 9 total pairs/systems on a live 1000 account... oh joy.
It is what it is, until its not, and then its something else.
Uczestnik z Sep 04, 2009
879 postów
Jan 19, 2010 at 20:41
Good luck!
Uczestnik z Sep 04, 2009
879 postów
Jan 25, 2010 at 03:49
account R.I.P.?😈
Uczestnik z Aug 31, 2009
131 postów
Jan 25, 2010 at 04:36
BOOM goes the Dynamite!!
It ain't easy being Cheesy!
forex_trader_7
Uczestnik z Aug 01, 2009
941 postów
Jan 25, 2010 at 05:32
(edytowane Jan 25, 2010 at 05:52)
I'm going to be the first person here to put down a losing system and leave it on....whooohooo!!
If a system wipes out leave it. Can learn as much from that as a winning one. Everybody has a few of those in their past. No shame in that. Can't learn to ride a bike without falling off. War scars are to be shown off!!
If a system wipes out leave it. Can learn as much from that as a winning one. Everybody has a few of those in their past. No shame in that. Can't learn to ride a bike without falling off. War scars are to be shown off!!
forex_trader_7
Uczestnik z Aug 01, 2009
941 postów
Jan 25, 2010 at 07:11
What I want to know is why do people call it a forward test when it is in fact a test. There is no such thing as a forward test.
A back test, tests the past. A tests, tests the present. A forward test would test into the future. Not possible.
A back test, tests the past. A tests, tests the present. A forward test would test into the future. Not possible.
Uczestnik z Aug 06, 2009
397 postów
Jan 25, 2010 at 11:10
Elkart posted:
What I want to know is why do people call it a forward test when it is in fact a test. There is no such thing as a forward test.
A back test, tests the past. A tests, tests the present. A forward test would test into the future. Not possible.
😁😁😁
i never even though about it until now..good point!
Sleep is for the weak.
Jan 25, 2010 at 16:42
I call it a forward test because its not a back test and its generally accepted that testing on a demo account with live data is a forward test, however, you have a good point.
Yes, the account died, went from 135% profit to 35% loss. I made some adjustments a couple weeks ago that proved to be a bad idea :(
I guess I could have left it up, but didnt see the point. I will be putting a new one up soon when I figure out what exactly about the changes I made were critically wrong. It actually ended up being just on the EURJPY and I think I know what I did wrong... the trigger to enter a trade was to easy to hit... in fact, looking at it, it was impossible not to hit, so there was a long series of loosing trades. However, this is why Im doing this, to identify issues before I go live.
Yes, the account died, went from 135% profit to 35% loss. I made some adjustments a couple weeks ago that proved to be a bad idea :(
I guess I could have left it up, but didnt see the point. I will be putting a new one up soon when I figure out what exactly about the changes I made were critically wrong. It actually ended up being just on the EURJPY and I think I know what I did wrong... the trigger to enter a trade was to easy to hit... in fact, looking at it, it was impossible not to hit, so there was a long series of loosing trades. However, this is why Im doing this, to identify issues before I go live.
It is what it is, until its not, and then its something else.
*Komercyjne wykorzystanie i spam są nieprawidłowe i mogą spowodować zamknięcie konta.
Wskazówka: opublikowanie adresu URL obrazu / YouTube automatycznie wstawi go do twojego postu!
Wskazówka: wpisz znak@, aby automatycznie wypełnić nazwę użytkownika uczestniczącego w tej dyskusji.
