logo
Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

Options
Go to last post Go to first unread
jjstein  
#1 Posted : Sunday, October 16, 2005 6:25:45 AM(UTC)
jjstein

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 5/13/2005(UTC)
Posts: 715
Location: Midwest, USA

Was thanked: 1 time(s) in 1 post(s)
After installing the 9.1 update, I re-ran some System Tests and got COMPLETELY different results than with 9.0! I did NOT change the System Test parameters in ANY way!!! Anyone have ANY idea what was changed in 9.1? Is this a bug in the new version? Did they fix something in the old version? What the HECK is going on!!!??? --Johnathan
Patrick  
#2 Posted : Sunday, October 16, 2005 3:43:48 PM(UTC)
Patrick

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 9/8/2004(UTC)
Posts: 2,266

Was thanked: 1 time(s) in 1 post(s)
Wow ... well yes that has to be a bug ... Let's tell everyone about the bug ... :roll: No need to make sure your options were the same, that you loaded the same amount of data ... I can guarantee you that your settings right now are not the same as they were before ... Comission settings, entry delays, slippage, equity % etc ... Patrick :mrgreen:
Marilyn  
#3 Posted : Sunday, October 16, 2005 4:05:17 PM(UTC)
Marilyn

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 9/10/2004(UTC)
Posts: 863
Location: Salt Lake City, UT

:) Be nice! What Patrick is trying to say is... (not completely positive but...) there really wasn't anything done from 9.0 tp 9.1 in the System Tester. It was more about connectivity to QuoteCenter, etc. Will you please double check all the settings again? Thanks - M What??? do ya need a coffee or something Patrick ? :D
hayseed  
#4 Posted : Sunday, October 16, 2005 9:56:11 PM(UTC)
hayseed

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 3/7/2005(UTC)
Posts: 1,346

hey johnathan..... was courious if you perhaps changed data vendors or reloaded historical data..... you didn't mention if the difference was across the board or just with 'csco'..... i myself really don't do alot of system testing but have ran across times were bad data could skew results..... if in the upgrade you reloaded historical data , the cummalative change over time due to correct data could be quite large...... in which everyone could be correct.... no change 9 vs 9.1 and no change in your settings ...... just a thought...... just a thought.....h
jjstein  
#5 Posted : Sunday, October 16, 2005 11:57:11 PM(UTC)
jjstein

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 5/13/2005(UTC)
Posts: 715
Location: Midwest, USA

Was thanked: 1 time(s) in 1 post(s)
PATRICK: I've thought of and checked most of what you mentioned -- Commissions, delays, etc. Also load options. Anything specific that I've missed, which can cause major differences? Have printed out system tests and checked params carefully; cannot find any discrepancy!!! MARILYN: I only got the upgrade because the salesman said that, outside of Quotecenter -- in which I have NO interest -- it was a bug fix version. Also, I talked to Robert at Tech Support, and he said they DID fix some things, but did know the specifics. He was supposed to get back to me with a list, but it's been several days (nearly a week) already. HAYSEED: No change in vendors, or reloading of data. Just installed 9.1 upgrade CD, cost a whole $4.00. Now, wish I hadn't -- don't know whether the current version is right, or the previous. And I'm going to guess that my charts and templates are now "stamped" with the 9.1 version code, and won't work if I try to go back. Anyone have any idea how hard it will be to install 9.0 over 9.1? --Johnathan
henry1224  
#6 Posted : Monday, October 17, 2005 12:22:48 AM(UTC)
henry1224

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 10/29/2004(UTC)
Posts: 1,394
Location: Glastonbury, CT

Was thanked: 2 time(s) in 2 post(s)
on your system, click edit on your General tab, order bias check long orders, Portfolio bias check single, Position Limit check mark and set limit to 1 on the buy order & buy to cover tab, order type check limit enter this Ref(2*((H+L+C)/3)-L,-1) in limit or stop price entry size should be default size expiration should be good till canceled delays should be set to 1 on the sell order & Sell short order tab, order type check limit enter this Ref(2*((H+L+C)/3)-H,-1) in limit or stop price entry size should be default size expiration should be good till canceled delays should be set to 1 under system testing options initial equity set to 10000 default size is number of units set to 100 Portfolio set to both, close all position on last bar click on the more button set your broker tab for margin and commission set your trade execution tab to realistic market prices report back with your results!
hayseed  
#7 Posted : Monday, October 17, 2005 3:10:49 AM(UTC)
hayseed

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 3/7/2005(UTC)
Posts: 1,346

hey johnathan..... another thought that has thrown me in the past was the unintentional inclusion of interest in the talley.... that option will of course change all results.... the reloading of meta 9 is nothing more than sliding in the orignal cd..... on most of my machines i have both 9 and 9.1.... its a somewhat simple matter to switch back and forth.... even though its confusing at this point, it might pay to stay in the fight a couple more rounds before you throw in the towel..... ya never know, the 9.1 version might be giving the correct results.....h
jjstein  
#8 Posted : Monday, October 17, 2005 1:01:30 PM(UTC)
jjstein

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 5/13/2005(UTC)
Posts: 715
Location: Midwest, USA

Was thanked: 1 time(s) in 1 post(s)
HENRY: Using your settings, the results are different than current ones, but not the same as the original. There should be 8 trades, 4 each long & short. Using "BOTH", there are 6 to 7 trades. Using "LONG", there are between 1 and 3 trades. Using "SHORT", there are between 3 and 4 trades. FYI: The test orders are quite simple:
Buy/Cover: cross(month()=10,0.5) Sell/Short: cross(month()=5 and dayofmonth()>=15,0.5) Test symbols are DIA, IWM, QQQQ and SPY Date range is 5/26/2000 to 5/28/2004 I also tried my ORIGINAL settings, which are: Test:
Delay is 0 (which should mean next day) Price is market No stops. Simulation:
75% of Equity ($14,000) Money Market = 0 Realistic prices I can save & attach my ORIGINAL and the CURRENT results page(s), if it would help. --Johnathan P.S. May I ask what is your logic for using a limit price of twice the typical, minus the low/high for buy/sell?
jjstein  
#9 Posted : Monday, October 17, 2005 1:32:34 PM(UTC)
jjstein

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 5/13/2005(UTC)
Posts: 715
Location: Midwest, USA

Was thanked: 1 time(s) in 1 post(s)
Going over printouts of the ORDERS page, I've just noticed that: On the OLD test (9.0 version):
The PRICE FIELD says "Open", which is as I understand it, should be correct if ORDER TYPE is set to MARKET. On the NEW test: (9.1 version):
The PRICE FIELD says "Calculated", even though ORDER TYPE is MARKET. Anyone have an explanation for this? --Johnathan
jjstein  
#10 Posted : Monday, October 17, 2005 2:28:16 PM(UTC)
jjstein

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 5/13/2005(UTC)
Posts: 715
Location: Midwest, USA

Was thanked: 1 time(s) in 1 post(s)
May have solved problem: Going over the SYSTEM page printouts, it seems that SLIPPAGE has changed from 0 to 2. I'm going to guess that the INSTALL did this, as I put it a 0 so as to get the same results from system variations. This is now producing the same results as the 9.0 version on the simple test. Hope to get the results on others. Thanks for the help, all! Although HENRY, I'd still like to know your logic for limit price.
henry1224  
#11 Posted : Tuesday, October 18, 2005 1:57:40 AM(UTC)
henry1224

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 10/29/2004(UTC)
Posts: 1,394
Location: Glastonbury, CT

Was thanked: 2 time(s) in 2 post(s)
If I get a signal then I use that Price for my entry or exit. A delay set to 0 uses the close of the signal bar A delay set to 1 uses the next bar, either market or limit price. I use a delay of 1 bar because the earliest that you can enter the market is the bar after the signal bar! Hope this makes sense. You download data and open your chart, you get a signal, make your decision to trade, the market opens on the next bar Buy:=2*((H+L+C)/3)-L; Sell:=2*((H+L+C)/3)-H; Buy;Sell; Plot it on a chart If you get a buy signal then enter at the buy price, if you get a sell signal then sell at the sell price, if you don't get filled just go to the next trade
henry1224  
#12 Posted : Tuesday, October 18, 2005 2:11:15 AM(UTC)
henry1224

Rank: Advanced Member

Groups: Registered, Registered Users, Subscribers
Joined: 10/29/2004(UTC)
Posts: 1,394
Location: Glastonbury, CT

Was thanked: 2 time(s) in 2 post(s)
Place these indicators on top of the price chart Bands Buy B1:=2*((H+L+C)/3)-L; B2:=((H+L+C)/3)+(H-L); B1;B2; Bands Sell S1:=2*((H+L+C)/3)-H; S2:=((H+L+C)/3)-(H-L); S1;S2; Use these values for tommorrow's entry price and protective stop
Users browsing this topic
Guest (Hidden)
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.