The max trade option is applicable to all open trades right? Is it possible to tweak it such that we can adjust the max trade for each currency pair instead? I think it would be alot better since there might be a case whereby:
You set max trades at 10
9 of them are gbp/usd while 1 of them is EUR/USD because the long/short signal is triggered earlier for the former. However, gbp/usd is heading for a loss while eur/usd is heading for a TP, which results in an unpleasant situation.
I think we might need tighter stop losses as I notice that winners are usually 20-30 pips while losers are around 100-150.
Hello TCO and others. Thanks for sharing an interesting EA and a nice job of coding. I have been runnung on $5000 demo for the past three days. 14 pairs. The first two days it knocked of two sets of winning trades each day, about 10 trades each time. 0.01 lots, 50 trades maximum. Yesterday, however was a diaster. At the moment, it has over 100 trades open all under water. Drawdown at the moment is -$1400. I would not be surprised if it closed these trades at a big loss today based upon a 1H signal change. The EA went long on the EURO pairs while the EUR is in free fall. The lesson for me at the moment is to reduce the max trades to perhaps 10. And something has to be done with the MACD change of signal stop. Not sure what. Going to let it play out and see where it goes. Margin level is a 179%.
I am still running version 2. Lets make sure I am doing it correctly, I have 1 hour charts opened on 4 pairs with MARI V2 running on those charts. I have all setting at default except max trades set to 10. Is that all that I need to do to use MARI?
Also should I move to V3 and do I run it exactly the same?
The max trade option is applicable to all open trades right? Is it possible to tweak it such that we can adjust the max trade for each currency pair instead? I think it would be alot better since there might be a case whereby:
You set max trades at 10
9 of them are gbp/usd while 1 of them is EUR/USD because the long/short signal is triggered earlier for the former. However, gbp/usd is heading for a loss while eur/usd is heading for a TP, which results in an unpleasant situation.
I think we might need tighter...
Ignored
Scroll back a page or 2. I asked that question already. Pleaes read thru the thread your answer is there.
Hello TCO and others. Thanks for sharing an interesting EA and a nice job of coding. I have been runnung on $5000 demo for the past three days. 14 pairs. The first two days it knocked of two sets of winning trades each day, about 10 trades each time. 0.01 lots, 50 trades maximum. Yesterday, however was a diaster. At the moment, it has over 100 trades open all under water. Drawdown at the moment is -$1400. I would not be surprised if it closed these trades at a big loss today based upon a 1H signal change. The EA went long on the EURO pairs while...
Ignored
When will traders first learn to understand manual trading? Idiots assume that EAs will work constantly regardless of market conditions and special days such as holidays, bank announcements, and major news releases. Try to figure out which one it is today.
Obnoxious2, now I understand your name -- you are obnoxious. been trading Forex manually for about three years now, with some success. Thought about not running this EA on the hoilday, but made a conscious decision to let it run just to see what it would do. That is what demo accounts are for. Just trying to make a contribution to the thread. Some people don't seem to be able to get the EA to open trades. I sure did not have that problem yesterday. So if I am an idiot to you, just don't read my posts.
Mav - assign a magic number to each of your pairs. Then the EA will only watch that pair. If you are trading five pairs and you set the number of trades to 5 the max trades you will have is 25 or 5 per each pair. Once you get 5 on one pair it will not open up any others for that pair but will still be looking to open trades on the other pairs.
My demo account took a nasty draw down yesterday and today. $500 demo account went down almost $200. I have 10 pairs trading 0.01 lot 5 trades max. I think I need to do some work on the SL settings as they seem to big and I'm concerned that a few of the trades went through them and didn't close. This big ding has me a little concerned to move it over to the live account just yet. I will probably put it on at least one pair just to make sure it works on my live account.
I'm also considering moving my signal generating time frame to M30 but keeping my trend filter at H1. I would also like to experiment a little with using signals from H1 and trend from H4. NB and SH do a lot of their trading on the H4 for more reliable signals.
Also, I believe the major reason why the EA takes so long to backtest is the print function to the screen. It has to do this constantly between every tick or bar analysis. We need to add a switch to the EA to turn the print off for backtesting.
Obnoxious2, now I understand your name -- you are obnoxious. been trading Forex manually for about three years now, with some success. Thought about not running this EA on the hoilday, but made a conscious decision to let it run just to see what it would do. That is what demo accounts are for. Just trying to make a contribution to the thread. Some people don't seem to be able to get the EA to open trades. I sure did not have that problem yesterday. So if I am an idiot to you, just don't read my posts.
Ignored
And you just confirmed your idioticy. If you really have been trading for 3 years then you SHOULD have expected crazy things to happen on a holiday. Think before you talk. I didn't see a useful thing in your post regarding the EA. Just sounded like some greedy person who wanted an "easy ride".
Mav - assign a magic number to each of your pairs. Then the EA will only watch that pair. If you are trading five pairs and you set the number of trades to 5 the max trades you will have is 25 or 5 per each pair. Once you get 5 on one pair it will not open up any others for that pair but will still be looking to open trades on the other pairs.
My demo account took a nasty draw down yesterday and today. $500 demo account went down almost $200. I have 10 pairs trading 0.01 lot 5 trades max. I think I need to do some work on the SL settings as they...
Ignored
EJ, the last external parameter in the properties box for the EA is "Show Settings". If you set it to false when you run the strategy tester, the print to screen is disabled. This speeds it up a little, but it still crawls along.
Scroll back a page or 2. I asked that question already. Pleaes read thru the thread your answer is there.
It's black friday, what do you expect? Markets are calm cause nobody in the US is working.
When will traders first learn to understand manual trading? Idiots assume that EAs will work constantly regardless of market conditions and special days such as holidays, bank announcements, and major news releases. Try to figure out which one it is today.
Ignored
ok! we get the point! keep a cool head, we'll see what happens on manic monday! mate.
"Try to figure out which one it is today."- Doh its friday - that weekend feeling
|
Membership Revoked
|
Joined Oct 2009
|1,495 Posts
Maybe there is a lesson here...
Fridays are always "crazy".
"Normal" fridays are like riding a crazy horse.
NFP (once a month) is like getting kicked by a crazy horse.
"Black friday" (once a year) is just getting another kick.
For all the "naive", rookies and the ones that just forget to turn the EA off:
Put an extern parameter in the EA: "Trade on Friday" and set the default to "false".
Please do not use this on a live account yhet. I have done a few backtests and found out some encouraging but also important information:
It has a very high win rate.
When it starts losing it can lose most of your account in a couple days.
Jan through Feb 2008 gained 60% in just 2 months. In March it loses almost the entire account. It still needs tweaking. I think the SL is too high and there might be other adjustments that can be made, but it is not ready for a live account.
|
Membership Revoked
|
Joined Oct 2009
|1,495 Posts
Well, if I see something that looks promising - I'm using it on live account.
Please - don't follow my lead...
From my humble ecperience - backtesting is only good for debugging.
Forward testing on DEMO account leads to only half the truth (sometime ago I was testing same EA simultaneusly on demo and live account - and guess what? different results!
So what I'm doing is testing only on live account - with every trade hedged by a special EA.
This way I'm only loosing the spread - I'm calling it a learning fee...
My results for this EA so far (since Nov. 23) is 6 loosing trades (not yet closed).
TCO, hope you had a great weekend. Spent some time trying to understand where the big losses come from time-to-time. You mentioned an MACD double dip, so I started there. The attached pic is from a Strategy Tester run from June 1, 2010 to the end of November. The pic is of the 1H timeframe, while the run is on the 5m timeframe. You can see that the EURUSD market was in a rise for all of September. According to the SellReady 1 rules, sells are permited everytime the above-zero MACD dips down below the red MACD signal line. I have annotated several sell trades that the Strategy Tester took with red vertical lines. There were actuall 10 sell trades initiated over those few bars marked with vertical lines. All 10 of those trades ended up taking 500 pip stop losses. This happened several times during the test period, destroying the P/L. We need to give some thought on how to correct this. Here are a couple of obvious options:
1. Simply employ one of the stop loss features available in the MARI EA;
2. Have the EA close all sell trades when the MACD moves back above the signal line. negating the validity of the SellReady1 signal. Likewise for Buys;
3. Have the EA close all sell trades when currentMACD > previousMACD.
3. Do both 1 and 2 or 1 and 3.
It would be beneficial if we could figure out a way to solve this on the 5m chart. There is so much time lag on the 1H chart. Every bar on the 1H chart is twelve bars on the 5m chart.It takes hours to sense the change in trend on the 1H chart. Anyway, my thoughts.
I agree here on the shorter time frame signals for turning off trades. We are using the long term signals to validate the trend and the short term signals for entering the trade. IMO we should be using those same short term signals to tell us it's time to bail. Even if the hourly trend is still correct, I don't like holding numerous positions through a retace and bounce back in the direction of the trend....if it happens. The trend could be reversing on us before the H1 time frame finally catches up. I think the EA should be allowed to turn of trades on the shorter time frame.
Also, I did some testing on higher time frames. If the EA is set to conservative mode, it makes about 4 trades (all winners) a year on H4. If you turn off RSi and zero line settings, it will make more like 12 trades but they are wild and mostly losers. I think I'm going to stick to the lower time frames, but it does look like the H1 setting is a good baseline for determining the higher TF trend with this system.
Sorry I haven't been on the thread at all this weekend. I have been sick with sore throat and ear infection and I just needed sometime away. Friday was not very good for the system. Big DD, but we will see how the EA does this week and recovers. I have been thinking about closing the trades like Easyrider suggested or maybe waiting for better confirmation of trades and not buying or selling until price goes beyond the middle TMA line, I don't know if that has any merit. I appreciate you guys testing it and coming up with suggestions. When it wins - it has a high winning percentage, we just need to figure out where the loss threshold is and take our lumps when we have to take our lumps. We will keep at it. Thanks!
All the loosing trades we are carrying are Eur based buys. Looks like the bailout of Ireland is complete, so the Eur pairs will probably take off at the open tonight. So they might all recover. Buit that still does not solve the systemic problem with the EA on large trends where 1H MACD corrects even though the trend is still up (or down). I have been testing some changes. Lets see if anybody can come up with an answer.
MARI - My Automated Revenue Invention EA
