Guru Review

Welcome Guest ( Log In | Register )

2 Pages V  < 1 2  
Reply to this topicStart new topic
> Backtesting pitfalls, Thoughts....
vitaminm
post Sep 1 2004, 11:07 PM
Post #11


Member


Group: Traders-Talk User
Posts: 4349
Joined: 9-December 03
Member No.: 681



calculation doesn't seem to be right

SELL SIGNAL : 20000516 1466.04 1387.77 1387.68 (984.90 : 2.05 : +205%)


--------------------
vitaminm
Go to the top of the page
 
+Quote Post
Lurk
post Sep 2 2004, 02:41 AM
Post #12


Member


Group: Traders-Talk User
Posts: 120
Joined: 12-March 04
From: Turks and Caicos Islands
Member No.: 1628



QUOTE (vitaminm @ Sep 1 2004, 11:07 PM)
calculation doesn't seem to be right

SELL SIGNAL :  20000516  1466.04  1387.77  1387.68  (984.90 : 2.05 : +205%)
*


Enter long position: 481.14
Exit long position: 1466.04

(1466.04 - 481.14) / 481.14 = 2.047 = 205% increase.

Seems correct to me? Do you get another result for this trade vitamin?

This post has been edited by Lurk: Sep 2 2004, 02:43 AM
Go to the top of the page
 
+Quote Post
vitaminm
post Sep 2 2004, 09:38 AM
Post #13


Member


Group: Traders-Talk User
Posts: 4349
Joined: 9-December 03
Member No.: 681



you are right.
show trades for this year.


--------------------
vitaminm
Go to the top of the page
 
+Quote Post
SideShowBob
post Sep 4 2004, 07:02 PM
Post #14


Member
Group Icon

Group: TT Member*
Posts: 1836
Joined: 9-September 03
Member No.: 32



Lurk,

I wonder if you do the same curve fit from the same start date but end it in 1996. I doubt you would get the same values for the moving averages. Do the values you get still give a decent profit in the range from 1996 to present? If not it's just a curve fit.....if they do work then it's a viable system, at least it should be....

SSB
Go to the top of the page
 
+Quote Post
Lurk
post Sep 4 2004, 08:06 PM
Post #15


Member


Group: Traders-Talk User
Posts: 120
Joined: 12-March 04
From: Turks and Caicos Islands
Member No.: 1628



Yes right conclusion, I dont think it's a viable or usefull system or anything like that, I belive it's just randomness that this particular MA pair comes out as the one yielding the best result. More of a experiment to see which SMA/ EMA pairs that gives the best profit yield for SP.

A good viable system should perform well on whatever you throw at it, be it a index, stock, commodity, currency, etc.

Here's top ranking for each category.
First colum is points profit and last one is number of trades.
CODE
SMA(x) SMA(x)

1949.26 SMA(6) SMA(8) 842
1973.98 SMA(223) SMA(216) 47
1980.33 SMA(225) SMA(211) 31
1990.13 SMA(5) SMA(11) 586
1990.48 SMA(7) SMA(8) 1046
2009.42 SMA(227) SMA(203) 25
2009.59 SMA(225) SMA(209) 31
2027.73 SMA(224) SMA(209) 31
2045.69 SMA(224) SMA(210) 31
2060.73 SMA(225) SMA(210) 31

SMA(x) EMA(x)

1717.66 SMA(253) EMA(194) 14
1718.29 SMA(254) EMA(197) 14
1720.48 SMA(256) EMA(200) 12
1720.55 SMA(257) EMA(202) 12
1721.27 SMA(256) EMA(201) 12
1724.88 SMA(257) EMA(203) 12
1730.63 SMA(254) EMA(196) 12
1731.26 SMA(255) EMA(198) 12
1732.39 SMA(255) EMA(199) 12
2491.55 SMA(3) EMA(4) 1476

EMA(x) EMA(x)

1473.16 EMA(204) EMA(66) 15
1473.74 EMA(222) EMA(14) 31
1474.52 EMA(198) EMA(20) 24
1475.68 EMA(271) EMA(1) 99
1475.68 EMA(272) EMA(1) 99
1478.81 EMA(202) EMA(15) 29
1480.76 EMA(228) EMA(55) 15
1483.69 EMA(199) EMA(16) 27
1488.93 EMA(209) EMA(18) 25
1488.93 EMA(210) EMA(18) 25

EMA(x) SMA(x)

1738.45 EMA(84) SMA(50) 58
1752.52 EMA(10) SMA(12) 618
1771.39 EMA(9) SMA(12) 582
1782.40 EMA(9) SMA(10) 744
1788.18 EMA(8) SMA(10) 688
1789.53 EMA(12) SMA(13) 580
1804.10 EMA(63) SMA(49) 115
1816.78 EMA(52) SMA(45) 149
1881.40 EMA(10) SMA(11) 688
1938.22 EMA(9) SMA(11) 652
Go to the top of the page
 
+Quote Post



2 Pages V  < 1 2
Reply to this topicStart new topic

 



RSS Lo-Fi Version Time is now: 3rd September 2010 - 12:08 AM
The Financial Ad Trader
The Financial Ad Trader