RE: Problem with TS 2000i tests

["Riccardo Ronco" <rronco@xxxxxxx>]

Here the code

Input: Brk(20), Capital(10000);

value4 = Int(Capital / close);

if close > Highest(High,Brk)[1] then buy value4 shares next bar at
market;
if close < Lowest(Low, Brk)[1] then ExitLong next bar at market;


As you can see nothing fancy!

Riccardo


-----Original Message-----
From: Jimmy Snowden [mailto:jhsnowden@xxxxxxx] 
Sent: 06 April 2004 14:32
To: Riccardo Ronco
Subject: Re: Problem with TS 2000i tests


Riccardo,

It could be somewhere else in your code but when I traded stocks I used
my own code to trade the number of shares

Input: Amt(10,000);
Var:NoShrs(0);

NoShrs=amt/close[1];

If this or that then buy NoShrs shares on the close;

Something like that.  It has been a long time so I don't remember the
exact code.

Best regards,
  Jimmy Snowden
mailto:jhsnowden@xxxxxxx


Tuesday, April 6, 2004, 4:40:10 AM, you wrote:

RR> Dear friends,

RR> I am testing a simple 10-day channel b/o model LONG only on daily 
RR> data.

RR> I bet on each trade $10,000.00 and the trade is executed at the next

RR> open.

RR> I am using the option page to set up the $10,000.00 to keep the code

RR> simple hence the decision to trade $10,000.00 is NOT in my code.

RR> On certain stocks it is all quite ok however on some stocks ALL the 
RR> trades, as seen on the list of trades, shows the right entry and 
RR> exit price but the single p/l is totally wrong for all the trades.

RR> I understand it could be the data but I can have one or 2 spikes and

RR> the errors should be limited in that specific data.

RR> Could it be a possible bug?

RR> Should I, instead, decide to write the code to tell how many stocks 
RR> to trade instead to allow Tradestation to do it?

RR> I really need your kind help

RR> Regards

RR> Riccardo




Outgoing mail scanned by Norton