Jiva Dragon

settings for EURGBP
Strategy Tester Report
JIVA Dragon EA DSS T3MA MAs testv11.0
FXDD-MT4 Live Server 5 (Build 226)

SymbolEURGBP (Euro vs. Great Britain Pound) Period1 Hour (H1) 2009.10.08 00:00 - 2010.10.01 23:00 (2009.10.08 - 2010.10.08) ModelEvery tick (the most precise method based on all available least timeframes)
Bars in test5611Ticks modelled7132335Modelling quality90.00% Mismatched charts errors0
Initial deposit10000.00
Total net profit9065.13Gross profit9333.11Gross loss-267.98 Profit factor34.83Expected payoff226.63
Absolute drawdown2396.67Maximal drawdown5507.76 (42.01%)Relative drawdown42.01% (5507.76)
Total trades40Short positions (won %)28 (60.71%)Long positions (won %)12 (66.67%)
Profit trades (% of total)25 (62.50%)Loss trades (% of total)15 (37.50%) Largestprofit trade1314.34loss trade-23.93 Averageprofit trade373.32loss trade-17.87 Maximumconsecutive wins (profit in money)5 (2415.51)consecutive losses (loss in money)1 (-23.93) Maximalconsecutive profit (count of wins)2415.51 (5)consecutive loss (count of losses)-23.93 (1) Averageconsecutive wins2consecutive losses1

My broker is FXDD, malta. As a US citizen my choices of broker are getting more and more limited. the test and settings are done with FXDD, malta. tweaking may be necessary for traders using other brokers.
I ran a test using the settings for eu from fxdd, and there were losses. I changed the channel value a few points and it is fine.
These are the settings for eu for Alpari, uk.

Strategy Tester Report
JIVA Dragon EA DSS T3MA MAs testv10.0
AlpariUK-Demo (Build 226)

SymbolEURUSD (Euro vs US Dollar) Period1 Hour (H1) 2009.09.09 00:00 - 2010.09.08 23:00 (2009.09.09 - 2010.09.09) ModelEvery tick (the most precise method based on all available least timeframes)
Bars in test5711Ticks modelled8123691Modelling quality90.00% Mismatched charts errors0
Initial deposit10000.00
Total net profit6587.40Gross profit6587.40Gross loss-0.00 Profit factor
Expected payoff253.36
Absolute drawdown1532.00Maximal drawdown2094.00 (17.45%)Relative drawdown18.62% (1937.00)
Total trades26Short positions (won %)12 (100.00%)Long positions (won %)14 (100.00%)
Profit trades (% of total)26 (100.00%)Loss trades (% of total)0 (0.00%) Largestprofit trade500.00loss trade-0.00 Averageprofit trade253.36loss trade-0.00 Maximumconsecutive wins (profit in money)26 (6587.40)consecutive losses (loss in money)0 (-0.00) Maximalconsecutive profit (count of wins)6587.40 (26)consecutive loss (count of losses)-0.00 (0) Averageconsecutive wins26consecutive losses0

seems like EA is ready for live now? didn't follow the thread for a long time now.... : /

short update please : ]

Yes, it looks like were are ready.
I'm sorry I am kind of strapped for time.
the thread is not very long.
If you take a minute or two to read through you'll catch the gist of things.
That's the great part about these threads. They are very good records of what has been happening.
I always keep the 1st page updated with the latest downloads etc..
Bast Wünsche

jiva...

ok , gonna try EA next week while sitting in the university : D
u r doing well with it live so far?

best regards and a nice weekend!

the stop loss is large on these settings.
In the past I have hesitated to use a large stop loss.
But with these settings the trend is so well defined and the trades are only with the trend much of the risk is filtered out.
Also these settings are for the 1 hour. So larger stops are not unreasonable.
I just feel more confident using, 140 and even 200 pip stops because the trend is so well defined.
That isn't to say there may not be a disaster at some point...
Maybe Taiykizx's Martingale hedge technique would be in order here. No stone unturned.
http://papelipe.no/var/weblog/storag...llerylarge.jpg

jiva,

thank you for many backtests.
I tested #342 setting, v11.0, Alpari UK/Live.


With #336 and #337, 2 losses(1 trade because ea takes 2 positions) with Alpari.
At the moment I cannot test gbpusd and eurgbp with high modelling quality....

ummmm.......
with jiva's mtf settings, H4 MA and D1 MA are used....
Those candles are much different if broker's timezone is different.
It causes different results by brokers....?
jiva, what do you think....?

Yes I considered that. For my own trading since US citizen my choices of brokers are getting narrower and narrower every day I just went ahead with 4h and day anyway for FXDD.
But it may not be all that significant, only in very rare cases. The ma is slower to change in the higher time frames. Gu uses 34 sma. the channel for all the settings is slow too, 55 ema. So maybe little tweak for different brokers. EU and EURGBP both use 10 sma but on the higher time frames it doesn't change that quickly. It may be the worst that will happen is some good trades are missed, since entry requires all time frames agree.

I think if there are losses they may come from different price feeds and the affect they have on the candles closing outside the channel on the 1hr. The candles on the 1hr form pretty much the same across brokers. So it is different price feeds not ma differences for trend indication on 4hr and day. As little as 1 pip difference can cause the candle to close outside the channel. Alpari channel for eu needed to be set at 65 rather 55.

Even Geoff's ea has little bit of differences for different brokers, works best for the broker he uses, Alpari.
It's going to be like that, not just the candle forming differently because of time frame but actually different price quotes since market is not standardized.
For my own trading I have to stay with FXDD.
They have made some "backroom" adjustments that completely circumvent US regulators new rules. So FIFO, no hedging, etc has no affect on usual way of trading. Can do just like before. No problem.

Using your mtf options has made the difference for the ea.
I didn't see it at the time when you surprised us with mtf, but it was sheer genius to add these. We can't thank you enough for your hard work. I look at all the work you have put into this ea and I am amazed. Ever grateful. Nothing short of miraculous.

I think a little tweaking here and there for broker differences in price quotes will get us by. I am extremely confident of this ea with the present strategy in place. Trend is everything and we've nailed it.

Maybe we want to see about Martingale hedge for emergency cases. It may work because the problem with Martingale is getting caught in a consolidation. But the bad entries with the present strategy using mtf, would come from the familiar buying a top and selling a bottom, which is usually a pretty quick and drastic turn around. So quick that the 1st target of 20 pips isn't even reached. Follow my reasoning? The bad entries I have seen are the top and bottom variety. Trend is so well defined in the strategy that consolidations are avoided. But it will take some more hands on to confirm that is the case. Because all it takes is one Martingale caught in a consolidation to "end the party".

Another reason I like using FXDD is because the spreads are fixed ( I don't mind paying a little more for a fixed spread) I can test, test, test, on the weekend.

Missed you.

I have a special need for the 20 sec delay, which you kindly added, but for most traders they will not need it. So just set it to 0. Notice, no pending orders!! mtf is that good... That is another reason I feel so confident about this strategy. The mtf defines trend so well there is no need even for pending orders!!

Better setting for EURGBP than previously posted.
Instead of 1st profit 20, 2nd profit 1000 trying to catch the big ones,
better overall results, is 1st profit 20, 2nd profit 50.
1hour

setting for EURCHF.
1hr

jiva please excuse another question ....
how many trades there are at each pair a week?
and how many trades there air at once max.? (margincalculation^^)


thank you so much for your work....ea does look pretty cool now

best regards and thanks in advance
PhanTi'

Looks like for EURUSD average about 3 or 4 month. So maybe 1/week.
Only 1 trade (2 positions) opened at a time.

Am going to spend time today studying the actual trades on the charts. since there are no losses I want to see how close the trade comes to being a loss and under what conditions it could happen. Because the stop is pretty big and a loss would be significant. But we don't have any losses to study

Оther result with presets on EURUSD tf=H4
and mistakes of the testing.

Dear Taiyakixz,
In studying the trades on the charts it seems that if we could find entries when all the available times frames agree we can get even better entries further minimizing drawdown and avoiding bad entries. Perhaps trading on a lower time frame.

So, question for you is: After this nice little vacation from coding would you consider adding 2 more ma options to the 5 we already have, so that we can set for 5 min & 1 min?

Is there any way to set for weekly in the tester? That is the really powerful trend indication.... Obviously, to live test forward for weekly will take a long time...

so 4 trades a pair means ~ 40 trades a month ( low spread pairs only).
so we have to push lotsizes to earn some nice money ....

how about trading a lower timeframe with more trades to spread risk over i.e 100 trades than over 10 trades while banking more but smaller wins?

The ea was set up in the tester but did not take any trades during the range bound market. That is very good. It only trades in well established trends. We like that very much.

Yes, that is the idea to be set up on several pairs.

I understand your point about lower time frames and I have put some time and effort into testing trades on the lower time frames, tweaking settings, etc., but this particular strategy does not produce the same results on other time frames. Somehow the 1 hr is optimal.

Reminder: this ea is extremely versatile and this strategy we are optimizing at present is just one of many possible strategies for the ea.

But it is by far the best one so far... for sure...

I will modify the EA to check 7 MAs and send it to you.
I don't know how to do backtest with weekly chart, but I think it's not so hard to do it manually. Even 10 years not so many trades will be done.