Ralph Ronnquist has worked with this...PM him..

ES

__________________
Live Portfolio (This takes a while to download as it is a big excel spreadsheet)

Thanks. Looking forward to his implementation.

<Right; I've done a bit like this before. I'll put this on the stack, but there are a few things before at the moment, so don't hold your breath....

regards,
Ralph.>

Please people dont post your journals here. Instead you could open a thread in the journal section. Maybe a mod could help keep this thread clean...

what timeframe are u using for carrygrid ea?

These are my bests..

Some tips for beginners:
---------------------------
-Focus on statistical/stochastic methods
-Neural etc. is not a proveable way.
-Focus on waves, not indicators
-Statistical models MUST include price itself, price itself again.
-Some pattern recognition/classification approach may help,
like classification of random trades mentioned in "better like" things.
-Focus on "analyzable" pairs.
-MT4 Tester: Strategy tester is a BIG power, if you can "glue" the price, when
your expert makes a decision. MT4 price simulation model papers will help..
-Count tester trades and real trades. They have to match %90+. IF they do not match,
you need to "glue" the price somehow, by including the price itself in your code.
-Do not cheat yourself saying forward test etc. Statistical prove on market data is a must.






A Classification Example, Using Weka. X,Y: Some parameters of sample trades.
If you can export your system's all possible trades outside of MT4,
you can "visualize" your system in the space, your brain will like this.

Blue: Good Class Red: Bad Class

Thanks
Cengiz, TR

Attached Images

classi.GIF (52.6 KB, 1164 views)

If you seriously think those are good results then you should probably read a bit about risk management..

edit: here is an example of what a good equity curve looks like. It's always important to focus on risk, not profits. If you keep your risk minimal, then you can always trade with bigger lots when your EA makes constant profit. (that graph is from 1989 to today with a very small lot size)

Attached Images

TesterGraph.gif (9.3 KB, 1162 views)

thanks. mines are from 2006.06.01
these pictures are not so simply produced

Forex/leverage is a mathematical/statictical montage market.
If you have enough statistical model and a reliable broker, risk is just nothing up to half of your money. (except friday closes and weekends) I could analyse euru/usd up to 1:50 leverage. These things are terribly hard.

The Key Is: More Leverage/Risk -> Less position carry time
Risk = F(carry time) + F(market continuity) + F(Broker Reliability)

No backtest results are any good at all. Only forward tests prove anything worthwhile.

finally
my goal was to make a wage
still working on it
Im ashamed to say all done with indicatorz

Attached Images

StrategyTester.gif (5.0 KB, 1027 views)

Attached Files

StrategyTester.htm (15.1 KB, 119 views)

The result of your tester looks good, but the modelling quality is n/a

__________________
VPS For Trader Only
duyduyfx-Automated Forex Review