Cost Averaging System - page 10
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
I think there is a small unintended bug in the EAs. If the computer is shut down for some reason and then started, it forgets the last number of lots it bought. We had a power spike last nite and now I can see that it started a series from 0.2 lots again, when it was supposed to start at 0.5. I will have to figure out a way for it to calculate the lots each time it opens a trade.
As for the system, yesterday was a great trend day. The EA is taking a beating, atleast the one I am running. (v3 on M1 charts). There are a bunch of open trades. This is a great stress test for the system.
Thankyou Maji for the version v2_1. It showed me that the timeclose method alone won't resolve the catrastrophic drawdown.
However,
It appears that these settings...
Parameters MagicNumber=12323; MA_length=10; applied_price=5; LotExponent=5; slip=3; Lots=0.1; TakeProfit=7; Stoploss=500; PipStep=86; TimePeriod=0; MaxTrades=2; UseTimeOut=false; MaxTradeOpenHours=25;
do resolve the catrastrophic drawdown, at least in the backtest on the USDJPY pair.
I would have posted the report from the demo account which has now reached $700 which is up from $500 initial deposit on the 11th when I started the tests on three pairs. Sadly something has changed the account and it's deleted all history and now seems inoperable so I'll have to start over.
p.s. yes I know the modeling quality is only 24% but I've never seen any 1m report that was any better???
Aaragorn,
The version v1 of the pyramiding EA did use the MA envelopes like you suggested. However, those were really not a good indicator for trend exhaustions. That is why I wanted to see how random entry works.
I think indicators like RSI, CCI etc are better for pointing out trend exhaustions and that is why I came up with the RSI based system. Anyway, these are all experiments and one has to do hundreds of them, before one finds a test which is successful, just like you have been doing.
Thanks again.What about ATR and Candlesticks? Those show trend exhaustion pretty well don't they?
I don't know what to make of this result. This is the gpbjpy, it appears to be similar to the usdjpy pair except it has one catastrophic drawdown on the very end....
why did it change the s/l to allow that on the very end when it had managed it well all the rest of the time?
Hello Aaragorn,
I am surprised that a Senior Member like you makes results of modeling with 24%
Hello Aaragorn, I am surprised that a Senior Member like you makes results of modeling with 24%
lol sad isn't it. doesn't say much for the backtester either...
tell ya what if you can tell me how to get better results on 1m data I'm all ears.
Hello Aaragorn, I am surprised that a Senior Member like you makes results of modeling with 24%
Without ticklevel data you cannot get better than 25% quality. Please refer to literature posted by Metaquotes where they explain their interpolation process.
What about ATR and Candlesticks? Those show trend exhaustion pretty well don't they?
ATR and candlesticks do not show trend exhaustion. Combined with other things they may show that, but by themselves they do not. You can find isolated examples of them showing it, but on a consistent basis they do not.
Forward testing
Cost Average v3 moving profitably but look at the explosive gains with RSI.
John
The first day forward testing Cost Average v4. Results are posted here.
John