Trading formula follow-up

[AlexPali]

Hello Panagiotis;

I’m still waiting for the answers or the D1-3 trading method that you requested.

Meanwhile, from Ignacio provided with a spreadsheet designed to automate set up calculations for the PM1 and PM2 (D1-1 & D1-2) methods. It simply requires keying in the requested data values from applicable previous days.  I hope the above-mentioned spreadsheet would be helpful tool during developing of Pip-Hunter. 
However, I found that the set-up output results calculated by Nacio’s spreadsheet differ from the ones computed by the existing Pip_Hunter beta version robot.
At this point, we’re not sure if Ignacio made some mistake in his Excel spreadsheet formula or Pip_Hunter has a glitch in it’s code algorithm.
Attached please find the FX_Journal_April-09.xls spreadsheet file and Trade-monitoring-chat_2.doc file containing the conversation with Nacio where I found out the differences of the two calculation methods.
The file contains 3 pages and I edited it by highlighting some text portions to make it easier to follow.

Regards,
Alex

Below is the chat conversation:

[/b]

------------------------------------------------------------------------------------------------------------------
Live trade-monitoring chat with Nacio        Wednesday, May 20, 2009

AlexPali says:
Good morning Nacio;
Hey I wish we are trading live money on GBP.USD!
It went up so nicely for the last two nights.

Ignacio says:
OOHH YES, PT2 HITTED
OR TRILING STOP ON THE 2ND LOT
POS CLOSED ON MY END +610 PIPS TOTAL

AlexPali says:
Very good 
Now we have to monitor all the trades and compare the positive against the negative ones to get the average.

Ignacio says:
YES, WE CUT EVERY MONTH SO WE COMPARE CURRENT MONTHS RESULT AGAINST THE PREVIOS

LOOKS LIKE USD/JPY PM2 SHORT WILL TRIGGER SOON, IN CASE U WANT TO GO FOR IT

AlexPali says:
I see you're making monthly statements. When did you start monitoring all trades you made? How about putting all of them into the same spreadsheet to average.
Yes, let try the Yen to see how it works in the long run.
This way we'll know if USD.JPY is potential pair to consider live trading down the road. Right now we are just speculating that it is dangerous but we don't know it for sure.

Ignacio says:
GOOD DEAL
I LIKE DOING MONTHLY CUTS FOR EVALUATION PURPOSES ITS JUST MY STYLE, BUT INDEED YOU CAN DO ONE WITH ALL TRANSACTIONS NP
Y, I AGREE

AlexPali says:
Well you can share all your previously monitored trades so I'll continue adding new ones.

Ignacio says:
LETS START FROM WHAT I ALREDY SENT APRIL-2009
IT COUNTS FROM WHEN WE'VE BEEN FOLLOWING UP TOGETHER THIS FX DEAL
THAT WAY EVERYTHING ITS FRESHH

WE ARE MISSING ONE ELEMENT ON THE SETUP CRITERIA ON THIS U/J TRADE
LOOKS WE DONT HAVE;
4.- LL3>LL10
PLEASE DOUBLE CHECK

I THINK THE U/CHF, DID TRIGGER SHORT A PM2 @ 1.1031
IN PREVIOUS BAR

AlexPali says:
The issue here is that the brokers have slightly different server time.
FXDD - 00:29  already started new D1 candle 0.5 hour ago.
while 
FXCM 22:29  will close the D1 candle in 1.5 hour from now.
Thus according to my calculations the set ups are as follows:,
with FXCM  - USD.JPY PM1 Short at 94.814  and USD.CHF PM2 Short  at 1.1058
While with FXDD: no PM1 and no PM2 for USD.JPY yet.
But there is a PM1 Short for USD.CHF at the price: 1.1014  on FXDD (server time: 00:46)

Ignacio says:
The issue is does it meet all criteria for a valid PM2 setup???
2pm CA time should be the day close
USD/JPY is not meeting PM1
since the blue & red MA's are going down for a while now

PM1 setup should have a change of direction of one or both red & blue MA's on the same direction from previous bar
PM2 would be;
hort trade
1.- THE 40 MA < 40 MA OF 5 BARS AGO
2.- HH10<HH35
3.- THE 3 MA (BASED ON THE HIGH) IS GOING UP ANY TIME IN THE PAST 5 BARS
4.- LL3>LL10
5.- ADX>16 ANY TIME IN THE PAST 15 BARS
6.- STAND ASIDE IF THERE IS EVIDENCE OF BULLISH DIVERGENCE AS DEFINED BY THE PIP REV METHOD

AlexPali says:
Please confirm the USD.CHF Nacio.

Ignacio says:
what was the low in your chart on 5/13/2009 bar?

AlexPali says: USD.CHF
5/13/2009 bar;     High: 1.1099     Low: 1.0977
5/14/2009 bar;     High: 1.1121     Low: 1.1025
5/15/2009 bar;     High: 1.1247     Low: 1.1037
5/18/2009 bar;     High: 1.1264     Low: 1.1138
5/19/2009 bar;     High: 1.1163     Low: 1.1069
5/20/2009 bar;     High: 1.1128     Low: 1.0968

Ignacio says:
Our charts differ a few pips +/- but should be no problem; bottom line should be the same deal
: we will have some differences on Entry, SL, PT levels
a few pips more or less

AlexPali says: That is fine as long as the set-ups are matching in general.

Ignacio says:
According to your info your entry/trigger short level should be @ 1.0958
According to mine 1.0962

(Pip_Hunter calculated D1-1  Short for USD.CHF at the price: 1.1014)
AlexPali says:
Hmm, I'm making mistake because I coming up with the 1.1014 price value.

Ignacio says:
how?
remember when looking for LL3, you should count from and including the setup bar low
In this case you want to look for Lowest Low of 5/18, 5/19, 5/20
in your case that would be 1.0968

AlexPali says:
We should developed an Excel spreadsheet enabling us to input the applicable values from the previous days and get the simulated set-up results.

Ignacio says:
Well actually the excel file I sent you does tell automatically the Entry, SL, PT, and PT2 levels
That is why before putting the order with broker you should fill out the excel format.

AlexPali says:
Oh right, I haven't used it yet. Thought it is just fro keeping record of the trades.

Ignacio says:
spotting the setups gets easier once you practice more and more
trading live would be a full time business my friend
that is why it takes time to switch

Ignacio says:
Alex I have to go to an appointment, will be back later

AlexPali says:
Thank you very much.

------------------------------------------------------------------------------------------------------------------


[attachment deleted by admin]

[AlexPali]

Folks;
Perhaps the differences in determining the set-up entry levels between Pip_Hunter and the FX-Journal.xls
occurs only when the calculated D1 bars are a combination of two weeks linkage versus calculation involving only Mandy trough Friday D1 candles?
This idea is raking my mind since in the past Pip_Huter’s computations were off just 2-4 pips versus Nacio’s manual calculations.
I wonder if there is a way to make the Excel spreadsheet to obtain the required daily indicator values automatically from reliable source such as MT4 feed for example?
I believe at least one of our members is dual; MT4 and Excel guru able to figure it out.
Regards,
Alex

[xux99]

I have coded the first beta version according to the rules of the first word document:

D1-2_Short

TRADING RULES, SHORT POSITION
1.- 40 MA of today < 40 MA of 5 bars ago
2.- 3 MA (based on the high) is facing up anytime in the past 5 bars
3.- LL3 > LL10
4.- ADX >16 in the past 15 bars

As I can see there are two additional rules not included in the EA:

HH10<HH35
and
STAND ASIDE IF THERE IS EVIDENCE OF BULLISH DIVERGENCE AS DEFINED BY THE PIP REV METHOD

This may explain the difference. Could you please post a screenshot of the USDCHF trade with the relevant indicators attached on chart? The EA did not open a D1-2 short order on my Alpari account. What do you mean by "combination of two weeks linkage?

Here you can find a video explaining how to import live MT4 data feed to excel. You can use:

=MT4|BID!EURUSD
=MT4|ASK!EURUSD
=MT4|HIGH!EURUSD
=MT4|LOW!EURUSD
=MT4|TIME!EURUSD

[AlexPali]

( …. Could you please post a screenshot of the USDCHF trade with the relevant indicators attached on chart? The EA did not open a D1-2 short order on my Alpari account. …)

Unfortunately, I can’t supply with the requested screen-shot because I used Pip_Hunter with unchecked “Allow live trading” due to the 6 digit prices (FXCM broker) yet not supported by the current version of Pip_Hunter. Thus, I’ve been using Pip_Hunter just as a guide not robot to trade FXCM demo manually.

(…What do you mean by "combination of two weeks linkage? …)

Perhaps the differences in determining the set-up entry levels between Pip_Hunter and the FX-Journal.xls occurs only when the calculated period of time starts during previous week and includes days of the current week when the set-up matures; versus calculation involving only Mandy trough Friday period of time of a single week?
Previously, (before Nacio developed the FX-Journal spreadsheet), Pip_Huter’s computations were off just 2-4 pips versus Nacio’s manual calculations. However I can’t recall if the previously matching calculations occurred during one or two consecutive weeks.
Is this explanation clear enough?

Here is another indicator setting details that could have influence on calculated values of set-up entry price levels.

FPA Indicators Settings;

IF CHARTING SOFTWARE DISPLAYS THE SUNDAY BAR USE;
Simple MAs_ 48 (Purple),
21 (Red),
11 (Blue),
3 (Based on the Low, Brown),
3 (Based on the High, Brown)
ADX (Average Directional Index)_ 28 (Yellow, Pane1)
ATR_ (Average True Range) 9_ (Red, Pane1)
Stochastic Slow_ 9, 3, 3,
Upper Level Line 85, Lower Level Line 15
(K% Red)
Price Lines_ Entry Level (Green),
PT1 (Blue Solid),
PT2 (Blue Dotted),
Current SL (Red Solid),
Previous SL Levels (Red Dotted),
Major Sup/Res Levels (Yellow)

IF CHARTING SOFTWARE DO NOT DISPLAY THE SUNDAY BAR USE;
Simple MAs_ 40 (Purple),
18 (Red),
9 (Blue),
3 (Based on the Low, Brown),
3 (Based on the High, Brown)
ADX (Average Directional Index)_ 14 (Yellow, Pane1)
ATR_ (Average True Range) 5_ (Red, Pane1)
Stochastic Slow_ 8, 3, 3,
Upper Level Line 85, Lower Level Line 15
(K% Red)
Price Lines_ Entry Level (Green),
PT1 (Blue Solid),
PT2 (Blue Dotted),
Current SL (Red Solid),
Previous SL Levels (Red Dotted),
Major Sup/Res Levels (Yellow)

[xux99]

Extra digits brokers are supported by the current version of Pip_Hunter. What problems did you encounter? If FXCM has a Sunday bar you should use different indicator settings and this may explain the differences in calculated entries.

[Tim]

Guys,

Is the EA posted on the forum?

Or are we waiting until the next beta version of the EA is ready before posting it?

Also is the EA going to contain all 3 strategies in the one EA? Or split into 3?

I think from what I saw in the very first draft it had all 3 built into the one EA. Is this the best approach, and would it not possibly be easier to split it into 3 so that each system can be thoroughly tested.

Cheers,
Tim

[xux99]

The first EA version had only the first two strategies and the rules were somewhat different from Ignacio's rules. By the end of the next week I shall post the next version with the three strategies. Since D1-2 strategy uses D1-3 strategy signals, it would be better to have all of them together. Each one can be traded independently and seperately though.

[AlexPali]

It looks like the existing Pip-Hunter is making very significant mistake while detecting D1-1 Long set-up conditions.
On Thursday, right after midnight GMT the robot detected D1-1 (PM1) Long set-up on EUR.USD as well as D1-2 (PM2) Long one within relatively short distance from each other;
D1-1 Long at: 1.3991 & D1-2 Long at: Long at 1.4036
The D1-2 Long happened to be a valid detection by the way. It kicked in and at the FX closing time on Friday there is 101 floating profit.

While consulting that situation with Ignacio, he was surprised that I’m speculating D1-1 condition since there was no sign of changing direction of Moving Average (9) and Moving Average (18) indicators; a must condition for D1-1 set-up in this case.

Thus, the portion of the code recognizing the above conditions probably should be revised.

[xux99]

The first version of the EA is based on different rules, that's why the signals differ.

[AlexPali]

Hello friends & Forex brothers;

Firs of all I would like to thank very much to Panagiotis (xux99) for delivering the following v.0.1 of Pip-Hunter robot. This is the very fist release incorporating all three main tactics of the strategy. (Still there is aggressive sub-D1-1 strategy to be added down the road.)

So far I noticed the following:
Since Pip-Hunter is using D1 Time Frame for analyzing the market movements, the result on spotting out a set-up is vary to brokers using different server times.
For example, FXDD has server time is same as Ignacio’s used FX Solutions but FXDD-UK and Alpari-UK are behind one and two hours. Therefore, the readings of D1 candles differ a lot among the three brokers and as a result, the set-ups determined by Pip-Hunter are different for hours ore even jumps to the following day if not disappears for a while.
Since we don’t know which starting/ending D1 Time-Frame is the most optimal for Pip-Hunter’s strategy, the only choice is to try all of the spectrum used by the most respected brokers and compare the final results.
In my first tasting days with two brokers: FXCM-UK and FXDD I noticed that the robot is placing the same trade twice in the case of FXDD-EUR/USD; (screenshot attached).
In the case of FXCM-UK the robot positioned D1-1 and D1-2 trades while the other one was still running.
I don’t know if such situation is permissible in theory and allowed in practice but this is something to be concern of.
I believe a coder and the only native Anglophone member among our team; Tim would be the best specialist to verify this condition against the available written strategy manual. (Well, Steve “forex5x” who is one of my best FX fiends is Anglophone too but he is sort of passive member so far).
Once again, many thanks for Panagiotis for his extraordinary job and for Tim for his further support in improving/tuning-up this robot.

With best Regards,
Alex

Quote:
AlexPali says:
It looks like all 7 USD involved pairs are heading for PM2 set-ups. But it might change over the night.

Ignacio says:
Remember you don't trade any PM1 or PM2 until PM3 has run its course.

"If a PM 1 or 2 setup occurs, but a Pip Rev trade is already in progress, then stand aside until the Pip Rev Trade has run its course."

[xux99]

I have posted v0.2 which should solve the double orders problem. Aggressive D1-1 strategy has been added. We need to confirm that Aggressive D1-1 strategy rules as described in the user guide agree with Ignacio's ideas.

[AlexPali]

Hello partners;

The last two weeks were not lucky for PipHunter Robot.
I’ve been running the EA on two; FXDD and FXCM-UK demos.
The two brokers have 2 hours server time differences so the entries are not matching due to the differences in daily candles values, an essential detail for PipHunter to determine entry levels.

I noticed that in average 70+% with both demos the entries went into floating profit but never reached the 1st or the 2nd Take-Profit levels before going into DD.
That shows the weakness side of the way the original PM strategy is handling open trades.
Perhaps the Break-Even feature that Panagiotis is going to add will stop the losing behavior of the system. However, there must be found another solution how to capture pips of the in-floating-profitable trade after it reverses the direction to the wrong way.
Attached are statements of the two FXDD and FXCM-UK demos.

With best wishes,
Alex

[xux99]

A new version with breakeven option and ZZ based exits has been posted.

[FreddyFX]

Have it loaded and running again.