Phreaky Phorex Successful Manual EA Needs Coding Experts - page 15

New comment
 

i like it david, lets test it out and implement it in the new EA if we can.............i want to help produce an EA that will shake the forex world!!!!!!!!keep on phreaking

 

Just an idea. Some of you senior coders may have tried this already. When the settings of the indicators are

extern int FastPeriod = 24;

extern int SlowPeriod = 20;

Instead of

extern int FastPeriod = 20;

extern int SlowPeriod = 25;

Just like the VTtrader.

You turn this EA into a guaranteed loser. (see graph)

Why not switch the buy/sell signals and trade the reversal? Then we code in another switch for the 100 pip runs. Thoughts?

Files:
testergraph_1.gif  8 kb
 

It looks like you lost $9318 over about 1700 trades.

Thats $5.38 per trade, or in other words less than you paid on the spread to make those trades ?

So, even reversed this EA makes money ... for the broker!

If so, if you reversed your buy/sell you'd really lose money.

Is this right, or am I late night crazy?

 
 
jarcrocker:

I also found an issue of Active Trader Magazine with a code for writing Chandes original indicator. It goes as follows..

For "Measuring Trend Momentum" by Tushar Chande, starting on p. 62:

{Tushar Chande 2001: VIDYA/CDMA}

We might be able to weed out false signals if we recode the VIDYA.

Hi,

You've found a real jewel. However, that article produced a bit of confusion in naming conventions. Or perhaps it "added to" the confusion. The code above, while bearing the word "Vidya" along with "CDMA" is actually CDMA, short for "Chande Derivative Moving Average". It's confusing that the word Vidya is in there, but they are similar but different indicators.

The actual "Vidya" Vidya code appears in Beyond Technical Analysis, by Chande, on page 36 in the 2nd edition.

There are actually several indicators floating around that are called Vidya / have Vidya in the name / or called Vidya_Something, that differ to various degrees from the published version.

For instance, John Ehlers, an absolute Moving Average Guru, published a slightly different one here: http://www.mesasoftware.com/Seminars/TSWorld05.ppt

(note there are several other interesting goodies there also)

Another item called "Vidya" that differs somewhat from Chandes' is here:

http://trader.online.pl/ELZ/t-i-VIDYA.html

There are several others out there.

However, the CDMA that you found is WELL worth studying on it's own merits. I have spent several hundred manhours with the Tradestation version of that bad boy, and variations thereof, just believe me when I tell ya. I don't know if it has characteristics that lend themselves to the Phreaky system, but it certainly has it's uses .

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

The Vidya code in Beyond Technical Analysis, by Chande:

{--Copyright 2000 Tushar Chande; Variable Index Dynamic Average VIDYA --}

vars: v1(0), v2(0), v3(0), v4(0), vidya(0), sfac(0.2) ;

{-- Calculate 20-day std. Dev. And its 20-day range --}

v1 = stddev(c, 20) ;

v2 = highest(v1, 20) ;

v3 = lowest(v1, 20) ;

{-- Create v4; stochastic oscillator for 20-day std. Dev.--}

{-- If v1=v2 (highest level) => v4=1; if v1=v3 (lowest level) => v4=0 --}

if (v2-v3) > 0 then v4 = ((v1 - v3)/(v2-v3)) Else v4 = 0 ;

{-- initialize and calculate VIDYA--}

{when V4 = 1, sfac*V4 = 0.20, corresponding to 9 day exponential moving avg--}

If currentbar <=50 then vidya = c ;

If currentbar > 50 then vidya = sfac*v4*close + (1-sfac*v4)*vidya[1] ;

{---Plot data ---}

If currentbar > 50 then begin

plot1(vidya, "vidya");

end;

{--- End of Code --}

Figure 1.25 VIDYA code for Omega Research Tradestation

 
Tross:
It looks like you lost $9318 over about 1700 trades.

Thats $5.38 per trade, or in other words less than you paid on the spread to make those trades ?

So, even reversed this EA makes money ... for the broker!

If so, if you reversed your buy/sell you'd really lose money.

Is this right, or am I late night crazy?

You are correct. The reason why most mechanical systems lose over the long run is due to transaction costs.

 

NFP & Weekends

Phreaky,

Will you open a 2 new positions crossing the other direction shortley before markets close on Friday, say 2-3 hours before close & then carry it over the weekend, or will you just close out in that instance & wait for new signal the next week?

Also do you try to prtect your profit before NFP or other volitale news? If reverse spike all of a sudden you can't get in at cross do you just get in as soon as you can and then take your 10 pips or do you wait 5-15 mins until market defines its direction?

Appreciate again your sharing. Thank you.

 

PhreakSystem - Maji Edition

Using Maji's, Vidya and Dema.

Trades are closing within 2 bars of line crossing.

Much better than within 4/5 bars.

Haven't seen any big swings since I set it running.

Losses are acceptable.

I'll let it run next week, unless someone posts an improved version that can anticipate crosses, tough to do I'd imagine!

 

Maji EA

Tross:
PhreakSystem - Maji Edition

Using Maji's, Vidya and Dema.

Trades are closing within 2 bars of line crossing.

Much better than within 4/5 bars.

Haven't seen any big swings since I set it running.

Losses are acceptable.

I'll let it run next week, unless someone posts an improved version that can anticipate crosses, tough to do I'd imagine!

Tross,

What are your settings for Maji indicators and EA ?

I am trying to test it but have some problems to synchronize croses with actual entrys.

 

One more thing about the EA

The Timeframe was externalize, I don't know why not just use Period() in the code. So, if you didn't change it to 5, the default is 0, which might be the problem for some people

tomstaufer:
Tross,

What are your settings for Maji indicators and EA ?

I am trying to test it but have some problems to synchronize croses with actual entrys.
1...891011121314151617181920212223
New comment